dna 2825f - dtic00 dna 2825f q4 final report december 1971 correlation of radar echoes from the...

00 DNA 2825Fq4 Final Report December 1971

CORRELATION OF RADAR ECHOES FROMTHE AURORA WITH SATELLITE-MEASUREDAURORAL PARTICLE PRECIPITATION

By: W. G. CHESNUT J. C. HODGES R. L. LEADABRAND

Prepared for:

DEFENSE NUCLEAR AGENCYWASHINGTON. D.C. 20301

CONTRACT DASA01-67-C-0066

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.

NATIONAL TECHNICALINFORMATION SERVICE

S-"AN Va 22153

STANFORD RESEARCH INSTITUTEMenlo Park, California 94025, U.S.A.

211

UNCLASSIFIED

DOCUMENT CONTROL DATA. R ________D___

Security *Iaassfacie,on of title. body' of Ab~terI .#eud I idexites! .mnnuggn ntust be .ftfr,. d wri 4, is,. t,.'.-r..U) rmp.rt ,. ti.,'.aed)

IORIGINA TING AC TIVITV (Corporatq uthor) 1 Z.C~4SECUF41TY CLASSIO-I'AT1,OI

Stanford Research Institute UnclassifiedMenlo Park,- California 94025 IZIC14U /

I REPORT TITLE

CORRELAT ION OF RADAR ECHOES FROM THE AURORA WITH SATELLITE-kEASURED AURORALN

PARTICLE PRECIPITATION4. DESCRIPTIVE NOTES (Tr'p of tepee: and inclusive dae&i)

Final Report Covering the period M~rch 1969 to December 1971L U THO4ORIS (First name. ahidcej injitia. last nameW

Walter G. Chesnut James C. Hodges Ray L. Leadabrand

6RPRDAE70. TOTAL NO OF PAGES Tb. tNo Or REFS

December 1971 121SOCONTRACT Oft GRANT NO *.OflIGI*4ATOORS REPORT Nu"BEIIIS)

Contract DASA-0l-67-C-0066 Final ReportNWET K11BAXH SRI Project 6548 :

- ~~~Task and Subtask X532 thais "oet rCS An fEtnmbr hIm) eesa

e. Work Order 01 DNA 2825F

Approved for public release; distribution unlimited.

III SUPPLEmEriTARY NOTES 12 SPOT.SORINGSMILE IARN AC TI VITY

Defense Nuclear Agency

jWashington., D.~C.**13 ABSTRACI

The characteristics of auroral particle prec~ipitation as measured during 14 passes bythe Lockheed OVl-18 satellite over Alaska have been compared writh. radar aurora simul-taneously detected at 139 and 398 Wiz beneath the satellite orbit. The radar aurorawere measured using the SRI auroral radar located at Homer in southern Alaska. Thesatellite measured precipi~ation of protons with energies above 10 keV and precipita-tion of electrons with energies above 800 eV. All measurements were made. nea.r localmidnight while the E.-region was in darkness.* The study has revealed the following:(1) Protons with energies above 10 keV regular~iy pr-ec-ipitate to the south of elec-

'rons with energies above 800 eV.(2) Electron precipitation in regions of radar aurora seems sufficiently intense toI

produce the E-region ionization needed to explain the observed radar scattering.I141*(3) Radar aurora (on 11 of 14 passes) was not associated with regions of intense pro-

ton precipitation.

(4) No obvious relationship existed between regions of radar atarora and regions ofmost intense flux of electron precipitation power (number f lax times nverage

electron energy), though there may be a texdency for these to be spatially

anticorrelated.(5) There seems to be a tendency for the elec-crons incideut into r'qgions of radar

aurora to have a slightly lower average energy (order of W80) than electronsprecipitaiitng elsewhere.

DD, "U6,47 E ASSIFIEDSIPE 0101-407.60O1 S9cuithvClassification

UNCLASSIFIEDSecurity Class-f;cation

I KEY WORDS A LINK, LINK C

ROLE *l ROLE WT ROLE WT

Aurora

Aurora correlation

Auroral radar

Satellite-measured auroral precipitation

V I

[I

t 4i

DD 1473 (BACK) -

(PAGE 2) securiky Classification

.:I .1-i. .. , _-

Item 13 ABSTRACT (continued)

(6) There is a tendency for the average electron precipitation power to

be less in regio-: of radar aurora than in other precipitation

There does not seem to be a deterministic relationship between the char-

acter of precipitating electrons and radar aurora--only subtle, statistical

tendenc.es were found.

I

Il

. I

Final Report DNA 2825FDecember 1971

'4

CORRELATION OF RADAR ECHOES FROMTHE AURORA WITH SATELLITE-MEASUREDAURORAL PARTICLE PRECIPITATION

By: W. G. CHESNUT J. C. HODGES R. L. LEADABRAND

Prepared for:

DEFENSE NUCLEAR AGENCYWASHINGTON, D.C. 20301

CONTRACT DASAOI-67--C-0066

This work sponsored by the Defense Nuclear Agency under NWET Subtask K11BAXHX532.

SRI Project 6548

Approved by: 45

DAVID A. JOHNSON. Directo:-

Radio Physics Laboratory

RAY L. LEADABRAND. Executive Director

Eiec.'vnics and Radio Scienae Divaion

CopY No ......... .

I.

ABSTRACT

The characteristics of auroral particle precipitation as measured

during 14 passes by the Lockheed OVl-18 satellite over Alaska have been4

compared with radar aurora simultaneously detected at 139 and 398 MHz

beneath the satellite orbit. The radar aurora were measured using the

SRI auroral radar located at Homer in southern Alaska. The satellite

measured precipitation of protons with energies above 10 keV and precipi-

tation of electrons with energies above 800 eV. All measurements were

made near local midnight while the E-region uas in darkness. The study

has revealed the fcllowing:

(1) Protons with energies above 10 keV regularly precipitate

to the south of electrons with energies above 800 eV.

(2) Electron precipitation in regions of radar aurora seems

sufficiently intense to produce the E-region ionization

needed to explain the observed radar scattering.

(3) Radar aurora (on II of 14 passes) was not associated with

regions of intense proton precipitation.

(4) No obvious relationship existed between regions of radar

aurora and regions of most intense flux of electron

precipitation power (number flux times average electron

energy), though there may be a tendency for these to be

spatially anticorrelated.

(5) There seems to be a tendency for the electrons incident

into regions of radar aurora to have a slightly lower

average energy (order of 89M) than electrons precipitating

elsewhere.

iii

(6) There is a tendency for the average electron precipitation

power to be less in regions of radar aurora than in other

precipitation regions.

There does not seem to be a deterministic relationship between the

character of precipitating electrons and radar aurora--only subtle,statistical tendencies were found.

iv

CONTENTS

ABSTRACT .. . . .. .. ... .. ...... ....... . . iii

LIST OF ILLUSTRATIONS ............ ..................... .... vii

LIST OF TABLES ............... ......................... ... xvii

AChNOULEDGMENTS .................... ......................... xix

1. INTRODUCTION ................... ..................... 1

2. EXPERIMENT GEOMETRY .............. .................. 5

2.1 Geometric Constraints on Radar Data ..... ....... 5

2.2 Geometric Constraints on Satellite Data .... ..... 6

2.3 Simultaneity of Radar and Satellite Measurements. 7

3. EQUIPMENT, OPERATION, AND SCHEDULE ............ ... 13

3.1 Homer Radar ....... ................... ... 13

3.2 OVl-18 Satellite ..... ................. .... 13

3.3 Operation and Schedule .................. ... 16

4. SATELLITE-RADAR-DATA COMPILATION ....... ........... 23

4.1 Maps of Alaska ...... .................. ... 24

4.2 Plots of Radar-Aurora Signal-Intensity .......... 26

4.3 Satellite Data ...... .................. ... 27

4.3.1 Electron-Precipitation Data ..... ........ 27

4.3.2 Proton-Precipitation Data ........ ......... 2S

4.3.3 Histograms of Electron-Precipitation

Data ........... ................ ... 31

4.4 Comments on Specific Satellite Passes ...... .. 32

4.4.1 S.tellite Pass A388 ............... ... 32

4.4.2 Satellite Pass A431 ............... .... 33

4.4.3 Satellite Passes A2979 and A2995 ..... .. 34

VA

I_ _ _ i

5. COMPARISONS OF RADAR AURORA WITH MAGNOMETER AND

ALL-SKY-CAMbERA DATA. .. .. ... .......................... 1075.1 Radar-Aurora Magnetomoter Comparison ..... ...... 107

5.2 Rada,-.Aurara-All-Sky-Camera Comparison .... .... 108

6. S• ".Z OF SATELLITE DATA ..... ............... ... 113

S6.1 E-Region Ionization Produced by PrecipitatingElectrons .......... .................... ... 113

6.2 L-Region Ionization Produced by Precipitating

Protons .......... .................... ... 116

6,3 CVrrelation of Radar Echoes with Satellite-Borne

Particle-Precipitation Measurements ....... ... 1186.3.1 Corrilation of Regions of Peak Energy

Flux with Locations Rada'." Aurora ...... 118

6.3.2 Preouction of Nighttime E-Region Ioniza-tion by Particle Precipitation ...... ... 125

6.4 Discussion of Correlation Study and OtherPossible Relationships. .............. 127

7. CONCLU3IONS. .......... ...................... .... 137

8. SUGGESTIONS FOR FURTheR WORK ............. 139

REFERENCES ............ ........................... ...... 141

DISTRIBUTION LIST ......... ........................ ... 143

DD Ferm 1473

vi

ILLUSTRAT IONS

Figure 2.1 Map of Alaska Showing Magnetic-Aspect Angles for

the Homer Radar Magnetic L Shells ...... ......... 8

Figure 2.2 Elevation View of Magnetic Geometry in Plane Con-

taining Homer and College, Alaska ...... ......... 9

Figure 2.3 Map of Alaska Showing Overlap of Radar and All-

Sky-Camera Coverage ..... ................ ... 10

Figure 2.4 Profile View Showing How Satellite Data and Radar

Aurora Data are Projected Along Magnetic Field

Lines to Ground Level for Purposes of Comparison . 11

Figure 3.1 Photograph of Humer Radar ............... .... 14

IPFigure 3.2 Block Diagram of 139-MHz and 398-MHz Radars .... 21

Figure 4.A.1 Map of Alaska Showing Ground Projection of Satel-

lite Orbit, Precipitating Electron Energy Flux

Along the Orbit, and Radar Auroral DistributionsDuring Pass A518 ....... ................. ... 35



Along the Orbit, and Radar Auroral DistributionsDuring Pass A641 ..... ................. ... 36



Along the Orbit, and Radar Auroral Distributions

During Pass A624 ..... ................. ... 37


lite Orbit, Precipitating Electron Fnergy Flux


During Pass A157 ....... ................. ... 38

vii

j

Figure 4.A.5 Map of Alaska Showing Ground Projection of

Satellite Orbit, Precipitating Electron EnergyFlux Along tne Orbit, and Radar Auroral Distri-

butions During Pass A388 ................ .... 39




During Pass A232..... .................... 40



Along the Orbit, and Radar Auroral DistributionsDuring Pass A676 ..... ................. .... 41

Figure 4.A.8 Map of Alaska Showing Ground Pro.)ection of Satel-lite Orbit, Precipitating Electron Energy Flux

Along the Orbit, and Radar Auroral DistributionsDuring Pass A751 ..... ................. .... 42


lite Orbit, Precipitating Electron Energy FluxAlong the Orbit, and Radar Auroral Distributions

During Pass A369 ...... ................ ... 43




During Pass A657 ..... ................. .... 44

Figure 4.A.il Map of Alaska Showing Ground Projection of Satel-


Along the Orbit. and Radar Auroral Distributions

During Pass A431 ...... ................ ... 45




During Pass A2964 ..... ................ ... 46

Figure 4.A.13 Map of Alaska Showing Ground Projection of Satel-lite Orbit, Precipitating Electron Energy Flux


During Pass A2995 ..... ................ ... 47

~31 viii




During Pass A2979 ....... ................ ... 48

Figure 4.B.0 Calibration of Radar A-Scope Traces for Use with

Figures 4.B.1 through 4.B.14 ............. .... 49

Figure 4.B.1 Radar A-Scope Records for Satellite Prss A518 . 50

Figure 4.B.2 Radar A-Scope Records for Satellite Pass A641 . . 51



Figure 4.B.5 Radar A-Scope Records for Satellite Pass A388 . 54


Figure 4.B.7 Radar A-Scope Records for Satellite Pass A676 56

Figure 4.B.8 Radar A-Scope Records for Stellite Pass A751 .. 57




Figure 4.B.12 Radar A-Scope Records for Satellite Pass A2964. . 61



Figure 4.C.1 Precipitating Electron Number Flux (TSUM),

Average Energy (SI), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A518 . . 64


Average Energy (W3N), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A641 . 65

ix


Average Energy (EMN), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A624 . . 66

Figure 4.C.4 Precipitating Electron Number Flux (TSUMI),

Average Energy (EMN), and Energy Flux (ELE)

Versus Universal Time for Satellite Pass A157 . . 67

Figure 4.C.5 Precipitating Electron Number Flux (TSUI),

Average Energy (EDN), and Energy Flux (ELE) Ver-

sus Universal Time for Satellite Pass A388. . .. 68

Figure 4.C.6 Precipitating Electron Number Flux (TSUBM),Average Energy (EMN), and 1ncrgy Flux (ELE)

Versus Universal Time ior Satellite Pass A232 . . 69

Figure 4.C.7 Precipitating Electron Number Flux (TSUM),Average Energy (EMN), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A676 .. 70

Figure 4.C.8 Precipitating Electroa Number Flux (TSUM),Average Energy (EMN), and Energy Flux (ELE)

Versus Universal Time for Satellite Pass A751 . 71


Average Energy (EBN), and Energy Flux (ELE)

Versus Universal Time for Satellite Pass A369 72



Versus Universal Time for Satellite Pass A657 . 73

Figure 4.C.11 Precipitating Electron Number Flux (TSUM),Average Energy (EM), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A431 . 74

Figure 4.C.12 Precipitating Electron Number Flux (TSUM),Average Energy (MIN), and Energy Flux (ELE)

Versus Universal Time for Satellite Pass A2964. 75

Figure 4.C.13 Precipitating Electron Number Flux (TSIMhI),

Average Energy (1201), and Energy Flux (EJS)Versus Universal Time for Satellite Pass A2995. 76

A



Versus Universal Time for Satellite Pass A2979. 77

Figure 4.D.1 Precipitating Proton Number Flux (B), Average

Energy (M), Statistical Error in Average Energy

(AE), Energy Flux (PAV), and Statistical Error

in Energy Flux (APAV) vs. Universai Time for

Satellite Pass A518 ........ ............... 78


Energy (E), Statistical Error in Average Energy

(BE), Energy Flux (PAV), and Statistical Error

in Energy Flux (APAV) vs. Universal Time for

Satellite Pass A641 .... ........... ...... 79


Energy (1), Statistical Error in Average Energy

I (AE), Energy Flux (PAV), and Statistical Error

in Energy Flux (APAV) vs. Universal Time for

Satellite Pass A624 ..... ............... ... 80



(AE). Energy Flux (PAV), and Statistical Error

in Energy Flux (6PAV) vs. Universal Time for A

Satellite Pass A157 ..... ............... ... 81




in Energy Flux (APAV) vs. Universal Time forSatellite Pass A388 ..... ............... ... 82

SFigure 4.D.6 Precipitating Proton Number Flux (B), AverageEnergy (E), Statistical Error in Average Energy

(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APWV) vs. Universal Time forSatellite Pass A232 ...... .............. ... 83

Figure 4.D.7 Precipitating Proton Number Flux (B), AverageEnergy (E), Statistical Error in Average Energy


in Energy Flux (APAV), vs. Universal Time forSSatellite Pass A676 ...... .............. . .. 84

I xi

MI

IN




in Energy Flux (APAV), vs. Universal Time for

Satellite Pass A751 .... ............... ... 85

Figure 4.D.9 Precipitating Proton.Number Flux (B), Average




Satellite Pass A369 ...... .......... .. ... 86





Satellite Pass A657 ..... .............. ... 87

Figure 4.D.11 Precipitating Proton Number Flux (B), AverageEnergy (E), Statistical Error in Average Energy



Satellite Pass A431 .... ............... ... 88


Energy (E), Statistical Error in Average Energy(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APAV), vs. Universal Time for

Satellite Pass A2964 .... ............... .... 89



(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APAV), vs. Universal Time for

Satellite Pass A2995 .... ............... .... 90

Figure 4.D.14 Precipitating Proton Number Flux (B). Average


(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APAV), vs. Universal Time for

Satellite Pass A2979 ..... ............... .... 91

Figure 4.E.1 Histograms of Logaritiniic Occurrence vs. Precipi-

tating Electron Number Flux (TSUM), Energy Flux(ELE), and Average Energy (EMN) for Satellite

Pass A518 .......... ................... ... 92

xii

I

Figure 4.E.2 Histograms of Logarithmic Occurrence vs. Precipi-

tating Electron Number Flux (TSUM), Energy Flux

(ELX), and Average Energy (DIN) for Satellite

Pass A641 ........ .................... ... 93


tating Electron Number Flux (TS'JI), Energy Flux

(ELE), and Average Energy (EMN) for Satellite

Pass A624 ........ .................... ... 94

Figure 4.E.4 Histograms of Logarithmic Occurrence vs. Precipi- Itating Electron Number Flux (TSUM), Energy Flux

(ELE). and Average Energy (EIM) for Satellite

Pass A157 ........ ................... ... 95



(ELE), and Average Energy (EWIN) for Satellite

Pass A388 ....... ................... .... 96

Figure 4.E.6 Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSUIM), Energy Flux

S(EIE), and Average Energy (MIN) for Satellite

Pass A232 . ................................... 97

Figure 4.E.7 Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSLTAI), Energy Flux

(ELE), and Average Energy (EIN) for SatellitePass A676 ...... ................. ..... 98

Figure 4.E.8 Histograms of Logarithmic Occurrence vs. Precipi- Atating Electron Number Flux (TSUM), Energy Flux

(ELE), and Average Energy (BN) for Satellite

Pass A751 ........ .................... ... 99


tating Electron Number Flux (TSUM), Energy Flux(ELE), and Average Energy (EMN) for Satellite

Pass A369 ........ ................... .... 100

Figure 4.E.1O Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSVID, Energy Flux

(ELE), and Average Energy (MIN) for SatelliteSPass A657 . . ..... .................. .. 101

xiii

13


tating Electron Number Flux (TSUM). Energy Flux


Pass A431 .. .. ....................................102

Figure 4.E.12 Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSUM), Energy Flux

(ELE). and Average Energy (EMN) for Satellite

Pass A2964 ....... .................... .... 103




Pass A2995 ....... .................... .... 104



(ELE), and Average Energy (MI) for SatellitePass A2979 ....... ....................... 105

Figure 5.1 Simultaneous Auroral Radar Echo Strength and

College, Alaska Magnetometer Deflections v.: Time

on 24 March 1969 ....... ............... . ... O110

Figure 5.2 Map Shouing Location of Visual Aurora and Radar

Aurora at 12:19 UT. 24 September 1969. Radarfrequency, 139 MHz ......... ............... 11

Figure 5.3 Map Shoulng Location of Visual Aurora and Radar

Aurora at 12:28 UT, 10 October 1969. Radar fre-

quency, 139 MHz ............ ................. 112

Figure 6.1 Altitude of the Maximum Rate of Production of

Ionization vs. Monoenergetic Electron Energy for

Monoenergetic Electron Spectrum. isotropic

angular distribution over dowmward hemisphere . 130

Figure 6.2 Altitude vs. Electron Density for Various Mono-

energetic Electron Fluxes. Electron density is

at ,ktitude of peak ionization rate ....... ... 131

Figure 6.3 Altitude vs. Chemical Time Constant for Mono-energetic Electron Fluxes. Time constant is

appropriate to altitude of peak ionization for

mongenergetic electrons ..... ............ ... 132

xiv

3_

- - -- ~-~-~---~~-- -zA

Figure 6.4 Altitude of Maximum Rate of Production of Ioniza-

tion vs. Mean Proton Energy for an Exponential IProton Spectrum, Isotropic Angular Distributionover Doumward Hemisphere .... ............ ..... 133

Figure 6.5 Electron Density at 109 km Altitude vs. Proton

Flux. Proton spectrum--exponential with E = 32

keV. Pitch-angle distribution is isotropic over

downward hemisphere ...... ................ ... 134

Figure 6.6 Average Properties of Precipitating Electrons vs.Satellite Pass Sequence Number Arranged According

to Ascending Three-Hour Magnetic K-Index .... ...... 135

Figure 6.7 Ratio of the Average Properties of Precipitating

Electrons in Regions where Radar Aurora wasObserve1', to those Properties in Regions where No

Radar Aurora was Observed, vs. Satellite-Pass

Sequence Number Arranged According to AscendingThree-Hour Magnetic K-Index ... ........... .... 136

ININ

xv

TABLES

Table 3.1 Homer Auroral Radar Parameters ... ........... ... 14

Table 3.2 Summpry of Characteristics of Satellite Passes

Used in this Study ....... ................ ... 17

Table 3.3 Simultaneity of Data Periods for Satellite Pass

and Radar Data ....... .................. ... 18

Table 4.1 Summary of Satellite Passes Ordered According to I

Magnetic Activity ....... .................... 25

Table 4.2 Summary of Definitions of Various Satellite Data

Quantities ....... ..................... .... 28

Table 6.1 Coincidences sod Anti-Coincidences Between Radar

Aurora and Peaks in Energy Precipitation from Satel-

lite Measurements ............ .................. 119

Table 6.2 Coincidences and Anti-Coincidences Between Radar

Aurora and Measurable Particle Precipitation 126 12

xvii

i7I

ACKNOWLEDGMENTS

We wish to thank Mr. Ronald Presnell of our Laboratory for the

assistance provided to this program in a number of ways. In particular,

Mr. Presnell has continued to modernize the Homer radar system and has

provided guidance on the operation of the instrument.

Miss Judy Moore provided computer programming necessary for the

reduction and presentation of the satellite data as it is found in this

report.

Dr. Richard Sharp and Dr. Richard Johnson of the Lockheed Space

Sciences Laboratory of Palo Alto, California, furnished the satellite

electron and proton precipitation data. We appreciate their guidance

and suggestions concerning the use of their data.

We appreciate the continued guidance and interest shown by Mr. Dow

Eelyn of the Defense Nuclear Agency.

The Geophysical Institute, College, Alaska, furnished all-sky

photographs.' The U.S. Coast and Geodetic Survey, College, Alaska,

furnished magnetograms.

The 1969 Radar Data Collection Program was supported both by the

Defense Nuclear Agency and by the National Science Foundation (Grant

GA-1061).

xix

I ___________ _____

1. INTRODUCTION

This report describes the research performed by the Rado Physics

Laboratory of Stanford Research Institute for the Defense Nuclear Agency

on Contract DASA01-67-C-0066. The program compares Alaskan auroral radar

echoes with the OVI-18 satellite data gathered by Lockheed Space Sciences

Laboratory.

The experiment is referred to as an input-output experiment. The

input to the ionosphere is measured by the satellite, and the resulting

effects (the output) are simultaneously measured by the Homer radar.

The program is a continuation of work started in 1965 under DNA sponsor-

ship. 1 The earlier experiment produced only one pass during which radar

aurora were present before the satellite orbit decayed.

S~The data used in this report were obtained under DNA sponsorship

between 24 March 1969 and 8 May 1969, and under NSF sponsorship from 3

September to 17 October 1969. These periods were near the maximum of the

11-year solar sunspot activity cycle. The data from 14 of approximately

60 satellite passes were made available for SRI by Lockheed late in March

& 1971. Five hours of magnetic data supplied by the Coast and Geodetic

Survey, College, Alaska, and five days of all-sky-camera data furnished

by the Geophysical institute of the University of Alaska are correlated

wi•th radar data in this report.

Although radar echoes have been observed from the aurora for approxi-

imately 20 years the mechanism that produces these radar echoes is stillRA

=*References are listed at the end of the report. S

ILI R

_ _ W:

It

not well understood. Earlier studies of radar data correlation with other

radio data, visual data, magnetic-field-fluctuation data, and drta from

other geophysical phenomena have not always shoun clear-cut relationships.

In particular, comparisons of visual aurora and radar aurora have showm

that the radar aurora is usually mucb broader geographically and usranly

occupies a somewhat different position in space than the visual aurora.•

Correlating satellite-measured particle precipitation with radar

aurora seems to be a fruitful technique to better investigate the under-

lying physics. The OVl-18 satellite that was used in this experiment

measured electron and proton precipitation at an altitude of 500 km over

Alaska. The Homer radar collected data magnetically below the satellite

during the time of the pass.

A report comparing precipitation during one daytime pass with radar

cchoes obtained at 1300 MHz by the Miillstone Hill radar has been reported

by Hagfors et al. 3 Their comparison also used the OVI-18 satellite. For

this one daytime pass they find that the radar aurora wa:; located in :,

region of proton precipitation. We find, and shall show later, that on

the night side radar aurora and proton precipitation tend to taLe place

at different latitudes.

This report is organized as follows. Section 2 discusses the experi-

ment geometry. The geometrical constraints on collection of radar data

and the geometric considerations relevant to radar/satellite comparisons

are presented.

Section 3 describes the characteristics of the Homer radar in suffi-

cient detail to understand this report. The characteristics of the

counters that were carried on the OVl-18 satellite are also described.

Section 3 also describes the operation of the equipment and the data-

collecting schedule.

2

i!j

Section 4 is a Cata compilation of satellite and radar data for 14

satellite past3s for which satellite particle-counter measurements were

provide%: to SRI by the Lockheed Corporation. The correlations that are

made have to do with the spatial location of radar and precipitating

particles.

Section 5 provides a comparison of some of our radar data with mag-

netometer measurements msde underneath the radar auroral scattering.I Section 5 alsc presents a comparison of the geometry of visual auroral

arcs with radar aurora obtained at the same time.

Section 6 is a study of the satellite data and correlations made

between the satellite and radar data in an atte:mpt to understand the nature

of the phenomena that are being observed. In particular, we concern

ourselves with production of ionization in the nighttime E-region, since

this ionization is necessary for the existence of radar scattering. We

also concern ourselves with the location of radar aurora and the character

of the particle precipitation.

Section 7 presents the conclusions that have been reached as a re-I

sult of this research effort.

Section 8 indicates additional work that should be done in order to 1

better understand the data that have already been collected.

3

IT

I

2. EXPERIMENT GEOMETRY

!IA variety of geometric considerations affect the interpretation of

the data correlations covered in this report. In this section we describe

some of these constraints so that the reader can view our data in better

perspective.

2.1 Geometry Constraints on Radar Data

As is well known., radar aurora is only obtained when the radar

beam line is nearly perpendicular to the earth's magnetic field in the

scattering volume. The signal strength depends on the magnetic aspect

angle; this angle is the difference between the angle exactly perpendicular

to the magnetic field and the actual angle between the radar beam line.

The Homer radar was located in order that this magnetic geometry constraint

could permit as wide a coverage of radar aurora over the state of Alaska

as possible. Figure 2.1 presents a plot of the magnetic aspect angle Tor

aurora scattering at en altitude of 110 km as observed from the Homer radar

site. Homer is located near the bottom of the map. As a practical matter,

radar auroras are usually detectable only When they occur within the 30

contour shown in Figure 2.1.

Figure 2.2 presents an elevation view. which is another way to

demonstrate the magnetic geometry. This view shows that the 00 magnetic -

aspect angle is a function of altitude when the radar views directly North.

Figure 2.3 presents a map of Alaska showing the 30 magnetic

ra aspect angle as seen from the Homer radar site for scattering at 110 kLmI

Figures are grouped at the end of each major section.

I'KLuk.,~jAf'PAGE BLANK 5ru

altitude. Also shown on the map is the area covered by an all-sky cameca

located at Fairbanks. The figure shows overlapping coverage. Another

all-sky camera, located et Ft. Yukon. extends photographic coverage

another 200 km north.

Figure 2.3 presents in R qualitative way the area of Alaska and

Canada that can be surveyed by the Homer radar in a practical sense.

During very strong radar auroras, echoes can be obtaii.ed from outside

this contour, but such reception is rare.

2.2 Geometric Constraints on Satellite Data

Figure 2.4 shows the way in which satellite precipitation data

were geometrically related to the radar data. This figure represents

an elevation view along a meridian through Homer and College, Alaska.

The figure shows various magnetic field lines. We show schematically how

satellite measurements of precipitation are projected down magnetic field

lines to the earth's surface. The radar aurora are also projected down

magnetic field lines to the earth's surface. Therefore, comparisons that

will be seen later will be between precipitation and radar aurora as both

are projected to the earth's surface.

There are three other geometry considerations concerning satel-

lite data that must be borne in mind, First, we project electron-

precipitation effects to the earth's surface as described by Figure 2.4.

!.- fact, ionization will be produced along magnetic field lines where the

electrons are precipitating. Therefore, there will be a real, geometrical

relationship in the radar scattering region, between electrons and radar

aurora. This is not the case for protons. Satellite-measured proton

precipitation is made at altitudes above 400 kilometers. Protons that

precipitate to lower altitudes where the atmosphere is more dense .. ill

suffer charge exchange. As a result, a very localized proton fluz at

6

satellite altitude nay be spatially spread by several hundred kilometers

by the time the protons reach altitudes between 100 and 11' kilometers.

SITherefore, relationships between regious of peak proton precipitation,

Sas measured by the satellite, and regions of radar aurora do not neces-

sarily relate to overlap of E-region ionization produced by the protons.

The relationship, if one does exist, is then between the mechanism that

causes proton dumping and that which causes spatial structuring of E-

region ionization that leads to radar aurora.

- There is a third geometry consideration regarding coi relations

I between radar aurora and precipitation. This relationship has to do

with the production of the E-region ionization that is necessary for

radar aurora. Since radar aurora most often occurs at an altitude on

the order of 110 kilometers. precipitating particles must have sufficiently

high energy to penetrate to this altitude if they are going to contribute

to the E-region ionization.

2.3 Simultaneity of Radar and Satellite Measurements

When we perform correlation studies between radar aurora and

satellite precipitation, it is essential, of zourse, that the measure-

ments be made at the same time. Since radar aurora can move large dis-

tances in a time of several tens of seconds, one surmises that the

simultaneity accuracy ought to be within 10 or 20 seconds. Pre-pass

ephemeris data were used in our analyses. Pre-pass ephemeris data was

sometimes in error. Four cases were particularly poor ia their simul-

_ •taneity. These four cases were passes A388, A431, A513, and A641. The

simultaneity for the last two cases uas in error by 7 and 15 minutes,

I respectively.

j Satellite precipitation data are output and recorded once per

second. During this time period the satellite moves about 7-1/2 km.

1 7

s1Satellite data are referenced to "satellite tir-r;." The post- -

pass satellite ephemeris is also related to this time. It is therefore

very important that the ephemeris geographical error be less than this;

otherwise our radar/satellite comparisons will not spatially overlap

properly. We assume Lockheed data are this accurate.

eoow

Lefr "L-7 1

LSHKELLSh - O0km .

S~.'• OFF I. ONTOURS.4h- 110 krn•;

AZIMUTH* L4STANUM PRIOJECTMO

FROM HOW"E

i FIGURE 2.1 MAP OF ALASKA SHOW'NG MAGNETIC-ASPECT ANGLESS~FOR THE HOMER RADAR MAGNETIC L SHELLS

I °aa

Aw

I5

Cm 42+I a

0 0

Fo

0 zUw z-i-

wC0 0

zz

o 0~

ItI

S-j

ww

z. I- Al

a:q

-. RADAR

ANDALL-KY-CAMEAMCOERAG

10

aa

CID uI

00LL<

CC.

0

ccwC

co-

11r

-- ~ ~ ~ ~ ~ ~ ~ a 0 -- =~-~ * ± _ -

ii--iI

S3. EQUIPMENI\, OPERATION, AND SCHEDULE

In this section we give pErtinent details of the instrumentation

Sthat was used on this program. \We also indicate the data-taking schedule

for the satellite passes for whih Lockheed has provided satellite data

for our use.

3.1 Homer Radar

The Homer Radar is located on Ohlson Mountain about 13 miles

from Homer, Alaska. Figure 3.1 is a photograph of the radar complex.

The latitude and longitude of the radar are as follows: latitude 59.7142*N;

longitude 151.5333 0 W. Table 3.1 provides a brief summary of the param-

eters of this radar. The six radar frequencies shoun in the table are

fed through the common 60-foot-diameter paraboloid. During satellite

passes all radar frequencies were turned on, were operating, and data

were recorded. In this work we have utilized only the data obtained at

139 and 398 MHz. A far more detailed distribution of the radar may be

obtained from a number of references. 2 Figure 3.2 provides a block

diagram of the 139- and 398-MHz radar systems as they were configured

for these experiments.

3.2 0Vl-18 SatelliteI

-he OVI-18 satellite was instrumented by the Space Sciences

Laboratory of the Lockheed Missile System Corporation. Details of the

satellite may be found in Lockheed reports.",s The satellite was launched

into nevr polar orbit. The satellite gathered 48 channels of data. The

data obtained by Lockheed were corrected for a variety of factors before

I

PREC0!Nl PAGE BLANK 13

If

0

LL

0

Ci

14-

II0 - 00 0 1 I

C0 0 0 t- cs0 .m $ ito 1-4 CMJ l

03 .4

gp t!. .0m~~ ~ 0 m C3 o m

5) I 4 C ) -4 C 4)

.44I)3ASI. 04

0)03 0)54) 0 5" 41 0 if)C

54)3 -44

.0 .

CiC3

*0 m 4 C

a 4J~ 4J. 4J a 0 -0 0 ' 44 ) 0 0. 0

C4) A. ) P. 54 <q ~ ~ .

15.

-~ -- - -~ -- -

being sent to SRI. For instance, counting rates were converted to ab-

solute particle flux; data were selected from many sets of counters accord-

ing to the momentary satellite orientation (the satellite stabilization

mechanisms failed to operate correctly); average electron energy was

computed, etc.

Basically, the satellite recorded electron-precipitation flux

over a range from 0.8 to 37.0 keV. These data were obtained by five

channeltron electron counters, each of which selected electrons over a

narrow energy range. Due to peculiarities of the satellite system, some-

times data were supplied from only the four lower-energy channeltron

counters. These data covered the energy range from 0.8 to 16.3 keV. The

electron-precipitation data were provided to SRI as TSUM or PSUM, which

means the total electron flux in electrons per square centimeter per

second per steradian. The TSUM included all 5 channels; the PSUM included

the flux in the lower four energy channels. The electron-precipitation

data supplied to SRI also contained the average electron energy obtained

by computing the first moment of the energy spectrum; this is obtained

by multiplying the energy of each counter times its particle flux in the

five (or four) channeltron counters. This quantity we label EXN. The

meaningful counting threshold for these counters was about 5 X 106 counts

per second.

The proton-flux data that were transmitted to SRI by Lockheed

gave the absolute proton fluxes as obtained by two proton counters. The

first is labeled the B counter. The B-type counters measure absolute

proton flux for protons with energies above approximately 10 keV. The

data provided to ShI were given in protons per centimeter per second per

5steradian. Meaningful counting-rate threshold was between 1 X 10 and

4 x 10 per second, depending on which of the B counters was being used.

Another type of proto,. counter was labeled the D counter. These proton

counters measured absolute flux of protons with energies above 38 keV.

16

The flux provided SRI was in units of protons per square centimeter per

steradian per second. Meaningful counting-rate thresholds were I x 105

to 4 X 105 counts per second, again depending on which counter was being

used.

A table presented in the data compilation of Section 4 provides

a summary of these counter characteristics, along with other data that

are pertinent to understanding the data presentations given there.

3.3 Operation and Schedule

The usual operational procedure was to receive satellite emph,!m-

eris data froin Lockheed for expected satellite passes. These ephemeris

data give predicted satellite position versus time. Accordingly, the

Homer radar would provide a sweep of the appropriate regions of the sky

during the satellite pass. Before and after the satellite pass the radar

would survey the entire sky over Alaska in order to observe the larger

picture of overall activity.

There were many occasions during which radar and satellite data

were obtained simultaneously. For reasons beyond our control, Lockheed

has provided data from only 14 of these satellite passes. These 14 passes

were selected because radar aurora or satellite precipitation was measur-

able. Table 3.2 presents a detailed summary of the 14 passes.

Radar data were obtained over a period of time, from considerably

before the expected satellite pass (usually several hours) to a time very

late after the pass was ex:pected to be completed (again, ui-ually several

hours). The radar data analyzed in this report are for the time period

determined from the pre-pass ephemeris data. As a result, the simultaneity

of data comparisons between radar and satellite is not always very close.

Table 3.3 summarizes this simultaneity more concisely than is obvious

*Table 4.2.

17

0wC -05 1at n C - t

t~! - N i lO C.0 IO N in0 2. 2

-0 -0C 0

0. 40 a &.

E-0 t- 0, In. =' : - - N N

W a . 0 E0 I 3.T z. 0 -

4. ft 34 ft 0 .o c

.. %,. E F- as. k 0 ;2. a . 0 0 ti a

tc to Wa 0~ a a2 w - 2.o a a Im A

A C

0 0 ) v 3 - 1) 2

S. 00 1.0 l ca:W. w C' q 4.r , ) t

. .4C 0 cr m 0 to v- ol No m~ 0to n 0.20 02.- ý m 0 4m NI N1 M Iv v %. 0 a2v. v-. v v L, 0 L) in L,) L3 to rnin - S. -

C4 in N.%n ý -

- 4 0m u

0 4

a4 00 0 0 - 06 N

0S 00 2D m 0 Nl 0 q 1- m 0 in 5.

a coo c! 0T!0 N -La------------------------- -4 0- -)e, -3 c C4 0 .

0 N ainI0A 0 0 " :; ! -at l l

0- - - - - -M -- - - - - - - - - - -- M eq C4 Q

CL in N u n in - N i

0 0 04 N . Ci No V - n4c - n N - 0 - - 0 i mi n 0 U

ino a..................................m m.O 00 0 0 0 0 0 - NM

- - - -- - -

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _41

-W a ltA:K

IMI

cs~

C3r - C') 10 -I D -4 C' 3 C) 4 k --

'-0

0:

C3

C) 4.4 CS

IDA 0 Id. 0 0 0 0 0D Le) 0 ) in IDQ0 4O ca oa3~ C) C ) 4 0

03 <

44 0 4)

0 c

cs E- C

EC 4 C' 0 0 00 to 00

0 04.. .=) C:

41-

4 to 0t- I-4C) C40 003 - 0 IT %0 t.0 03 03 000 4 CS) 03 cm M IV 0 0 W 0 w 0 to W - N* N 3%

41

j 19

from Table 3.2. Simultaneity of passes A431, A518, and A641 was exceed-

ingly poor. For passes A518 and A641 the poor time simultaneity probably

does not matter since our experience shows that the low K indices imply, geographically stable phenomena. Pass A431 occurred during high activity,

so that lack of time simultaneity of the measurements makes our compari-

sons less valuable for this pass.

The data collection was near midnight local time in all cases

except in October 1969. During October 1969 coordinated measurements

were made on eight midday satellite passes. No radar aurora was observed

during any of these daytime passes. We do not know if the Lockheed

satellite recorded precipitation in these regions of the ionosphere during

these pnsses.

The radar data that were obtained were tape recorded on a 14-

channel analog instrument. The data were read back in analog fashion.

The radar data were measured, geometrically corrected, and plotted on

maps that will be presented later, in Section 4. The satellite data were

reduced by computer before Lockheed delivered the data to SRI. At SRI

a number of computations were made on the satellite data. The results

of these computations are presented in Section 4.

Z

20

- --- -~-- -

zL~ui

2 cc

I- co

wIw

CC 0 c0 x

w Cx 0 c

Lu L-j

0. -j-i up

0Ou w I )

U- ý- r-);o

cc~ Lu c -L 20

OD z tr21

coIto Zw tca 04

I It1 4. SATELLITE RADAR-DATA COMPILATION

Satellite precipitation data were provided to SRI for 14 nighttime

satellite passes during which measurable precipitation and radar aurora

were simultaneously present. Note that radar aurora was not present

when there was no precipitation. Also, recall that during October 1969

I coordinated measurements were made on eight midday satellite passes.

SDuring these daytime passes, no radar aurora was observed, though at this

writing we do not know whether the satellite observed any precipitation.

I• The data obtained are presented in this section in several ways.

The main effort is to provide information that allows for a comparative

study of the spatial locations of radar aurora and the simultaneously

Sobserved satellite precipitation measurements. There are five kinds ofjdata presentations. They will be described in more detail below. Briefly,

they are as follows:

* The first format is a map of Alaska with the projection of Ithe satellite orbit down the earth's magnetic field lines

onto the map. Also shown on the map are the areas in which

radar aurora was found. The maps give an indication of

where precipitation was occurring.

* The second data presentation is a series of A-scope photo-

graphs giving radar aurora intersity.

* The third and fourth formats provide various data concerning

electron precipitation and proton precipitation versus

orbital time.

PRECEDING PAGE BLANK23

IN

.The fifth format provides histograms of occurrence of

electron-number flux, electron-energy flux, and average

electron energy for each of the 14 passes according to

whether radar aurora was or was not present in the precipi-

tation region. For easy identification, the figures in

these five formats are designated by figure numbers containing

the letters A, B, C, D, and E, respectively.

All data are labeled according to satellite orbit number. Data are

presented in order of ascending three-houi magnetic index at College,

Alaska. Table 3.2 provides a schedule of satellite passes and other perti-

at information. The times shown on all figures are Universal Time

(Greenwich Meridian Time). Subtract 10 hours to convert to Alaska Standard

time. Thus it will be seen that all measurements reported here were made

between midnight and 2 AM local time in Alaska. Note that because the

Homer radar can observe aurora over a wide azimuth around magnetic north,

radar aurora may be measured from areas that are experiencing significantly

earlier or later local magnetic time tnan at the radar site itself.

Dr-in all measurements reported here the E-region was in darkness. Table

4. 1 sts satellite passes according to ascending K index. The table

summarizes other pertinent information.

4.1 Maps of Alaska

In Figures 4.A.1 through 4.A.14 we present 14 maps of Alaska,

one for each of 14 satellite passes. The maps show the location of the

Homer radar facility near -he bottom of each figure. Each map is labeled

with the date of the satellite pass, the satellite orbit number, and the

radar frequency for which clutter returns are shown. The satellite orbit

as projected down the earth's magnetic field lines to the earth's surface

is shown on the map. Tic marks indicate Universal Time along this orbi-

tal projection. The total energy precipitated as a function of orbital

24

I•

Table 4.1

SUWMARY OF SATELLITE PASSES ORDERED ACCORDINGTO MAGNETIC ACTIVITY

Subjective Character

Time Simultaneity of Precipitation

3-Hour of Radar and (See Section 6

Sequence Pass Magnetic Satellite Data for description)

No. No. Index-K (minutes:seconds) Electrons Protons

I A518 0 7:00 Poor ....

2 A641 0 14:25 Very poor -- none

3 A624 1 0:00 Good ....4 A157 1 0:15 Good ....

5 A388 1 2:20 Good E --

6 A232 2 0:30 Good E wp

7 A676 2 0:10 Good E --

8 A751 3 0:35 Good E --

9 A369 4 1:45 Fair E --

10 A657 5 0:15 Good E WP

11 A431 5 3:20 Poor E P

12 A2964 6 0:45 Fair E none

13 A2995 7 0:15 Good E P

14 A2979 7 0:10 Good E PII

position along the orbit is shown by means of cross-hatched are:... The

amplitude of the cross-hatching from the satellite orbit is linearly

proportional to energy deposition. In an effort to make the presentation

more legible, the cross-hatched areas are extended on both sides of the

orbit.

Radar aurora are shown on the map by speckled areas. The re-

gion from Which clutter was obtained, if it was well mapped by the radar,

is also outlined with a heavy line. In other cases data are presented

25

--- M

only along the satellite orbit. In these cases the radar aurora is shown

as speckled with heavy boundaries that are not closed.

On several satellite passes t',ere are special considerations.

These will be noted on the figuizq.

4.2 Radar-Aurora Signal-Intensity Plots

Radar auroral echo strength and location of radar echoes are

determined in this report from A-scope photographic records. An A-scopa

record presents signal intensity or signal voltage as a function of rauge

from the radar. The A-scope photographs presented are time-exposure

photographs of approximately 30 seconds duration. The figures are anno-

tated with radar frequency used, time duration of signal integration

(Universal Time--subtract 10 hours for Alaska Standard time.), and antenna

azimuth and elevation. The A-scope records pro-iide approximate logarithm

of signal amplitude (post-detection amplitude) versus range in kilometers.

On some figures we have placed the symbol ;, which indicates that during

the integration time the radar aurora amplitude seemed to vary in ampli-

tude in a significant way.

Figure 4.B.0 presents calibration signal amplitudes that may be

used with the A-scope photographs. The calibrations are given in terms

of power input into the receivers for both the 139- and 398-MHz radars.

Calibration must be done in this way rather than as radar cross section

because cross-section calibration would be a function of range, aurora

geometry, antenna beam pattern, etc. A value of -100 dBm corresponds

approximately to 1I m2/m . This relationship assumes radar aurora is

on the order of 10 kilometers in vertical thickness, at a range of about

800 kilometers, and is beam-filling in the azimuth direction. Figures

are labeled according to three-hour magnetic index in ascending order.

The radar signal amplitudes for 14 satellite passes are presented in

Figures 4.B.1 through 4.B.14.

26

_ _ _ I4.3 Satellite Data

The satellite data are presented in the form of computer-drawn

graphs. Data are presented for electron precipitation and proton precipi-

tation separately.

4.3.1 Electron-Precipitation Data

Electron-precipitation data versus orbital time (Univer-

sal Time--subtract 10 hours for Alaska Standard time) are presented in

Figures 4.C.1 through 4.C.14. In each figure are shown plots of three

separate quantities. The top part of each figure presents T-PSUM. The

ordinate in the top figure is logarithmic in electrons per square centi-

meter per second per steradian. In regions where no data points are

shoum, the counting rates were low; at the suggestion of Lockheed person-

nel we have imposed a threshold at a value near to or just equal to the

on-board, radioactive-calibration-source counting rate. In some cases,,

through oversight, the amount of time for which data were supplied to us

was less than we required.

The middle part of each figure presents as ordinate the

quantity MIN. This quantity is the mean energy of the precipitating

electrons as determined by the five (or four) energy-selecting electron

counters. The units are linear from 0 to 10 kV. The bottom part of each

figure -resents 3 its ordinate the logarithm of the product of the T-

PSUM times EMN. ,e units are electron-keV per square centimeter per

second per steradian. This curve provides a measure of energy flux (power)

as a function of position along the satellite orbit. We have indicated

on the figures, by cross-hatched regions along the trajectory, where

radar aurora simultaneously existed. Table 4.2 provides a summary de-

scription of the quantities given in Figures 4.C.1 through 4.C.14.

27

27

Table 4.2

SUMMARY OF DEFINITIONSOF VARIOUS SATELLITE DATA QUANTITIES

k IElectronsI TSUM Total electron flux between 0.8 and 37 keV as measured by

five energy-selecting channeltron counters. MeaningfulSi ~threshold: 5.106 counts/s. Units: el/cm2-s-ster.

PSUM Total electron flux between 0.8 and 16.3 keV as measured

by four energy-selecting channeltron counters. Meaningfulthreshold: 5.106 counts/s. Units: el/cm2 -

EMN Average electron energy determined from counting-rate dis-

tribution in five (four) channeltron counters.

ELE Total electron energy flux (power density): (TSUM X EMN,

or PSUM X EMN).

Protons

2_B Proton flux, protons/cm -s-ster. Approximately l0-keV

threshold, no upper limit. Meaningful counting-rate

threshold: 1 X 105 to 4 X 105 counts/s, depending on thecounter.

D Proton flux, protons/cm2 -s-ster, 38-keV threshold, noupper limit. Meaningful counting-rate threshold: 1 X 105to 4 X 105 counts/s, depending on the counter.

EAverage proton energy under the assumption that the proton

energy spectrum is exponential. This quantity is deter-

mined from the relative counting rates in the B and D

counters.

AE Approximate fractional statistical error in proton averageenergy

PAV Total proton energy flux (power density)--(B counter fluxrate times E).

APAV Approximate fractional statistical error in total energyflux.

282$

IAs an aid in data-correlation work, an arrow has been

placed at the bottom of these figures. A vertical line carries the po-la

sition of the arrow across all of the subfigures. The arrow shows the

point in time when the satellite crossed the L-shell that goes through

College, Alaska. Also indicated is a distance in kilometers--east or

west of College--where this transit took place. The 3-hour magnetic K

index from the College magnetic observatory of the U.S. C(Anstal and Geo-

detic Survey is also indicated.

All satellite paths for which data are used in this re-

port were north-to-south transits. We have indicated on the figures that

the left-hand side is north and the right-hand side is south as a reminder

of this fact.

4.3.2 Proton-Precipitation Data

Figures 4.D.1 through 4.D.14 present proton-precipitation

data as a function of orbital position. There are five subfigures in

these figure sets. The top subfigure provides as ordinate the logarithm

of the proton flux in the B counters. The B counters provide flux mea-

surements for protons with energies greater than 10 keV. The units are

protons per square centimeter per second per steradian. Regions in which

there are no data plotted are regions where the counting rate dropped to

a value to be expected from radioactive calibration sources alone. At

Lockheed's suggestion we have imposed a threshole belou; which no data

points were plotted.

The second subfigure from the top provides as ordinate

the logarithm of the average proton energy versus orbital position. The

ordinate is labeled E. The average energy is computed in kilovolts. The

average proton energy is estimated using the assumption that the proton-

energy spectrum is exponential in energy. Under this assumption the

29

relative counting rates in counters Band D can be used to estimate theI

I average energy. Note that for an exponential energy distribution the

average energy and the e-folding energy are the same.

The B counters and the D counters are independent of one

another. Each counter, measuring proton flux over a short time period,

provides statistically independent samples. Therefore, each value is

subject to statistical fluctuations; when the proton mean energy becomesI

large, or the absolute counting rate becomes small, the determination of

the average proton energy, E, becomes statistically uncertain. Based onI

information provided to us by Lockheed, we have produced in the third

subfigure an estimate of the statistical accuracy of the E estimate. In

fact, the way data are handled on board the satellite does not allow for

good statistical estimation of data accuracy; it is believed though that

A• provides an approximate fractional, statistical error in our estimate

of proton average energy. This is a linear scale of fractional error

from 0 to 1. Certainly wh~en our estimate of this error is close to one,

the estimate of • is quite uncertain. When the estimated error is nearI

0, then our estimate of average proton energy is liable to be .•tatisti-

cally accurate; howe•ver, the reader must remember that the derivation of

E assumes an exponential form for the proton spectrum. We do not know

S~whether this analytical form is a good fit to the real precipitation

!a spectrum.

! The fourth subfigure from the top presents the logarithm

! of the total proton energy flux as ordinate. The ordinate is labeled

!t PAV. This quantity, PAV, is the flux in Counter B times the average pro-

S~ton energy; hopefully, PAV is proportional to the incident power density

S~of protons impinging on the top side of the ionosphere.

The estimatc of PAV includes the statistically uncertain

estimate of E. We have therefore provided in the bottom subfigure an

S~30 iI

1-

Iestimate of the fractional error in the total proton energy flux. This

fractional, statistic error varies linearly from 0 to 1. The ordinate

is labeled APAV. As with our estimate of the fractional error in E,

when our estimate of the fractional error in PAV is large, our estimate

of PAV is clearly very poor. When the fractional error, APAV, is near

0. the accuracy of the statistical estimation of PAV will be quite good;

however, th, reader must remain cognizant of the fact that the quantity

PAV includes an assumption of an exponential proton energy spectrum.

As indicated above, we do not know how valid this assumption is. Table

4.2 summarizes the quantities in Figures 4.D.1 through 4.D.14.

As with the electron plots, the location of the satellite

transit of the College, Alask- L-shell is indicated, as is the College

three-hour magnetic index. Satellite transit is from north to south.

This is also indicated on the figure.

4.3.3 Histograms of Electron-Precipitation Data

In order to seek a definitive correlation between radar

aurora and electron precipitation a series of histograms has been prepared.

Data have been organized so that two histograms appear c each subfigure.

Each histogram presents the number of measurements of a particular quan-

tity from regions of the satellite pass where no radar echoes were ob-

served (i.e., regions where they could have been observed it the environment

had given rise to scattering). On tne same subfigure another histogram

presents the same quantity for regions where radar echoes were observed.

The histogram sets are given in Figures 4.E.1 through 4.E.14.

The top subfigure gives two histograms of the electron

|nuber flux--TSUM--according to presence or absence of radar aurora.

Note that both the abscissa and the histogram interval widths are loga-

Srithmic. For gqeater ease of display the ordinate is also logarithmic--

with a place piiovided for zero cases. Satellite data are read once per

31 *

- i: N

second. Therefore the number of cases plotted corresponds to the number

of seconds of data that were obtained, subject to other constraints. The

particle-flux threshold value of 5 X 106 was applied in order to eliminate

data dominated by the radioactive calibration source.

The middle subfigure gives two histograms of the

electron-energy flux (ELE) according to presence or absence of radar

aurora. Note that the abscissa and ordinate sre Jogarithmic.

The bottom subfigure gives two histograms of average

electron energy (EMN) according to presence or absence of radar aurora.

Only the ordinate is Logarithmic in this plot.

4.4 Comments on Specific Satellite Passes

In this subsection we discuss features of several of the satel-

lite passes that may help the reader to better understand the limitations

of this experiment. Some of these comments relate to work that is pre-

sented later, in Section 6 of this report.

4.4.1. Satellite Pass A388

If electron precipitation in the area of Bar 1 (see

Figure 4.A.5) produces radar clutter, then it probably would not be seen

by the radar since the usual clutter altitude of 110 km is just beyond

the radar horizon. Furth-rmore, the magnetic aspect is 30Y further dis-

criminating against observing clutter. The area of weak clutter shown is

bounded in the north by the southern boundary of detectable electron pre-

cipitation (threshold taken to be 5 X 106 el/cm2 -s-ster). There is no

evidence of proton precipitation in the area. An undetected electron

6precipitation flux in this area with a flux just below 5 X 10 could pro-

4 3duce 6 X 10 el/cm . This density might be enough to allow auroral radar

echoes to be obtained should this ionization become structured into

irregularities.

32

4.4.2 Satellite Pass A431

The maps of Pass A431 (Figure 4.A.11) show that the ob-

served radar clutter does not coincide with regions of maximum energy

deposition. The plots of the B-counter (proton) and the TSUM (electron)

count show that, in regions where clutter is observed, there is precipi-

tation of both protons and electrons. The E-re-ion %as not illuminated

by the sun. A monoenergetic electron beam of 3 or 4 keV, the average

energy indicated by EMN, would not penetrate to the appropriate altitude.

We presume that the actual spectrum is spread in energy. The proton

average energy exceeded 10 keV in this same area of the sky; this average

energy can produce ionization in the relevant regions of the ionosphere.

The electron flux in the main clutter region varies be-

tween 10 and 3 X 10 el/cm -s-ster. The proton flux is approximately

3 x 10 . The data of Figures 6.2 and 6.5 (see Section 6) suggest produc-

4 5tion of steady-state E-region densities of from 8 x 10 to 1.5 X 10 due

4to electron precipitation, and - 8 X 10 due to proton ionization. The

steady-state electron density due to both electrons and protons would5

therefore be likely to exceed 10 This background ionization is probably

sufficient to produce observable auroral radar clutter.

We note that the electron-precipitation flux is very

much greater, by a factor of 10 to 30, in regions where no clutter isA

obtained, though time simultaneity of measurement is not very good. Most

observers believe that the electron-density irregularities that cause

radar scattering from E-region ionization are caused by electric fields.

This single satellite pass may suggest that regions of higher precipita-

tion levels produce higher conductivities that cause "short-circuiting"

of the ionospheric electric fields.

The region of intense electron precipitation between 640

and 650 north latitude lies in an area where the magnetic aspect angle

33

from the radar exceeds 3". Therefore, the radar cannot well monitor the

clutter environment in that region of the sky.

4.4.3 Satellite Passes A2979 and A2995

These two passes occurred during a high-energy proton

aurora. The proton flux into the B counters was not extraordinarily

large, being in the neighborhood of 10 protons/cm -s-ster. But the

energies of the protons were so large that the determination of average

energy--assuming an exponential spectrum--was jeopardized by statistical

errors. That is, the B- and D-counter counting rates were nearly equal;

this means that most protons had energies above the 38-keV threshold of

the D counters. The ratio of the q to D counting rate would be nearly

1.0. The logarithm of this ratio is divided into a constant to determine

E. Any statistical fluctuation in counting rate of either counter could

bring the ratio to 1.0, thereby driving the logarithm to zero and causing

us to compute an infinite average energy. The proton energies are so

high that an "average energy" determination from the counting ratio of

B to D is statistically meaningless.

Electron precipitation was taking place nearly everywhere

on Pass A2979. On Pass A2995 the electron precipitation was also wide-

spread. The average electron energy was high on Pass A2979, being about

6 keV. On Pass A2995 the electron energy was somewhat less.

Radar aurora wus intense and very widespread during

Pass A2979. The clutter was weaker and contained spatial gaps on Pass

A2995. We could characterize both events as occurring during intense

aurora, though A2979 was stronger. In spite of the apparent tendency of

regions of peak precipitation intensity and radar aurora to anti-correlateiN(not overlap) during satellite passes when auroral activity is weaker,

there seems to be good overlap in the locations of the two phenomena for

these two passes.

34

Li SATELLITE 7 MIUE

rG

FIGURE 4.A.1 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. AND

¥- ~RADAR AURORAL DISTRIBUTIONS DURING PASS A518

35%

I-g

I-li74-2 1PA5 159694

69.36

16 t

I&20

.1:0

WA4

A 4

RR~

FIGURE 4.A.3 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A624

37Iko

- ~ ~ ~ 0 :00~----

A 1A

A157

~2R.

cr KII

AZO~uTN-01STAME PROJECTON

IFIGURE 4.A-4 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT. -

PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT, AND 3RADAR AURORAL DISTRIBUTIONS DURING PASS A157

38

70.o

I AZAJTM-~StN~ POACT

I ~fi0:14:00FIGURE 4.A.5 MAP OF ALASKA SHOWING ROUDPOETINOtAELIEOBT

RADAR AURORAL DISTRIBUTION D~IN PS A

PRCPTTN6LCTO NRYFU6LOGTEOBT N

At" WEA

13 39Z

10160

74'a Is. lw S

TrIA~inJh-~SU~ M~CTM

10:51:00,

FIGUE 4..6 AP O ALSKA HOWIG GOUNDPROECTIN O SATLLT ORBIT

RADA AUORALDISRIBUION DURNG ASSA232

139 MI40 4ao

l~370 --rr IfAUOAI~RPS

0:38:0Io

10390IA

NO AURORA ON 139OR 398 MHz AT SATELLITE TIME

aZwmj-os. ~TAIcZ PROACTOVrnma .Dit

FIGURE 4.A-7 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PREC.PITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A676

41

AtI

|I

£ZE 6 MAY 1969--

11040 &ISM• 398 MHz

0 ono Nma

FIGURE 4.A.8 MAP OF AI.ASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT, ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A751

42

aA

_~1270 10-:0.!F

~~0A

RADA AURRAL ISTRBUTINS DRINGPAS 1396Este DoteýBlaire Lj'keA36

43!,"',13 ~

- ~ ~-----ELECTRON-

I

•*1

,10O-51:OO0I a II

10:53:00 •

S\30 APRIL 1969

1 e

.Z.-

FIGURE 4.A,10 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON 4.NERGY FLUX ALONG THE ORBIT, ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A657

44

( __

rT

10:08:0

18AIiI

RAARAUOALDITRB TINDUNGPS A431 7,

/f ___"'.

'A' R- -

IVIVIULAURORA -

12.21L-0

GrI

go.-'ArdI

FT. YKON 3

4696

%NON 39___M__z___

12:30:00

'.1.awl

r4*w

Al

\1 OCTOBER 1969

PRECPITAINGELECRONENERY FLX AONG HE RBICT.AN

RADAR AURORAL DISTRIBUTIONS DURING PASS A2995

47

Go- 4 -5

VIISUAL

A

NN

I 30 SEPTEMBER 1969

1-30 -139 Mhlz (398 ~ is ffmuaha)

404?

FIGURE 4.A.14 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A2979

48

; .~

-110 -100 -JO

-144-19

FIGURE 4.8.0 CALIBRATION OF RADAR A-SCOPE TRACES FOR USE WITH

FIGURES 4.B.1 THROUGH 4.B.14

3949

a.

00

In

to U)

F4 w 00 0j

00

4: 0

co 0C) LL

>D U, >

4- u

0 Im tN 0 0n 0

o c 0r< 0

> C.) w.

LLa

2

c~,50

oP

L i !N

C-i~

4 ~ ~ 00

> CUoU

0T <

0 0

z

0 0 0

>w

-j CP

I- s 0

4U cn

0 0

In On

z4

- 0

z B mI4F M

UJK w

I::C.,

51

L 0~

-

- 0i

0*--

N<

00 LU- U

zz z c

o2 "

0. cm 0 r 0 U

Q - r

I. C w

LU LUC-i -i U,LLJ L0

cc)

4 0 I- KV 00

UJ 52

FF-ý-

0 IaLn (D

I- Cr

an 0

00

InU

C~4 0.

n0~

I Ln

C4 0

I.-

4n --A

IRD

N <Ioolat DO

53

13 M(z 139

L111'LI< AI

SRANGE: -- km 144-18

•. ~12 APRIL 1969 -

"•, • FIGURE 4.B.5 RADAR A-SCOPE RECORDS FOR SATELLITE PAS;S A388

454

AZIMUTHi~rs2rr..o

_ _ _ _____ _ _

SAZIMUTH 3 ELEVATION 3' AZIMUTH 30 ELEVATION 3' AZIMUTH 3" ELEVATION 3'/4'

38139 MHz 139Mz

~Ifl

1050 00-105"30 1050.00-10500 10500-1051:0

Si I I t I I II I I I I

AZIMUTH 3' ELEVATION 4' AZIMUTH 3' ELEVATION 4' AZIMUTH 3 ELEVATION 4l

ELEVATION 4 AZIMUTH 3 ELEVATION 44 d

S139 M2Az 13913

A 0 00 50 050 5 10 50 05010 W

FZIUTH3'ELVTO 4 RAARI A-SCOPE ECORDS TFOR S^ATELLITE PAS ELEVTIO 4'

I0H~139 139MH1 55

IIA I.-

.8.

C-l,

ww

o So0 C0

Ca

LU)

LU cc0

z cc

00 ~C.

L <,

-~ C3I- C<

0 4U-j ccC-, I

Z *L

4??mo

o 70

3anil-dM 90

56

--------- .....

"AZIMUTH "V ELEVATION 2' AZIMUTH "•, ELEVATION 2' AZIMUTH 14' ELEVATION 'V

398 M~~z 398 MHz 38~

1100-5G-1 101:20 11301020 1020- 1102-'nSIII I I I I III

AZIMUTH 12 ELEVATION - AZIMUTH 12' ELEVATION 2' AZIMUTH "- ELEVATION 2'

0~4

S1103:00-110:30 03"30-1104 001102"30-1103:00

I I I I I I I I I

AZIMUTH - ELEVATION 2 AZIMUTH 'V ELEVATION 2' AZIMUTH ELEVATION 2 7'

1104:50111135.1104 1105 75 1105 55

tI I I I I i "I I I II

0 500 1000 1500 0 500 1000 1500 0 500 1000 1500

RANGE - -km R o14d-96 MAY 1969

FIGURE 4.B.8 RADAR A-SCOPE RECORDS FOR SATELLITE PASS A751

57

5 -

LUU- CD

C., -- 0

A o-J 0C L

ww

z 0

0 0

uj D

4 04

oc- CCI0

3a ild' eC. 74'

U) 0 S58

j; j0 C

o o c04 C) C3

Cd -,

C--J

on 8 tov

2 2ow 0

* w~ w

z 0

0 0 oa

> IU2i E

w 4 *Uz' <

0i0CIO

DD 0

0 - 0

-~~~~ NJOl Ii I ~

59l

I_<

>0 >

I- 2o

La w o

<I

('4 .7

-

4 4 0

0c 0 u

I-

2i x

0 ct ~o

-J LU ~

C.).

Ml Z

600

....... ... ... ... ......

-M127OD127:0-E12730-11228:0

E1228:00-1228:'o E1228.30-1229-01

J I II III IF0 500 1000 1500 0 500 1000 1500

RANGE - km 146-23I29 SEPTEMBER 1969

FIGURE 4.13.12 RADAR A-SCOPE RECORDS FOR SATELLITE PASS A2964

61

AA

IW 15:0-15 1154'30-1155'00

="1155-00-1155-: 1 0ý

IN -i 1

I OCTOBER I 96S T

000000m 1500:O-1543 150301150'0 0

RAGE -- k-14-

C% FIGURE 4.1.13 RADAR A-SCOPFz 13CORS FOR SATE-LLiTE PASS A2995

in

-

U-U

1211 112

1212 00-12123 1212 30-1213.00

0 Soo 1000 1500 0 500 1000 1500

RANGE - kni 147-8

30 SEPTEMBER 1969

FIGURE 4.13.14 RADAR A-SCOPE RECORDS FOR SATELLITE PASS A2979

63

,. ~ ~ w rW -- - -

I T 1010

T-P

SUM

106 . _ _ _ _

e I-1 0 .. ."

0 I

1010

ELE

106 A518

NORTH 10.14:00 10:15:00 10'16:00 10:17:00 SOUTHI 260 km -- E

FIGURE 4.C.1 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN)o AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A518

64

:1 ___

1010

1116

10

EMN

0

1010

ELEA'

BEYOND106 HORIZON __ _ _ A641

NORTH 11:05:00 11:06:00 11:07:00 11:08.03 SOUTH

240 km WK -0

FIGURE 4.C.2 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN), AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A641

65

Iý

1010

T-P

106o

10 I

EMN

ELE A

106 A624

NrJORTH 10:13 00 1r.14:fl0 10:15.C" 10:1C 00 SOUTH

3r.0 kr .- EK-I

F;GURE 4.C.3 PRECIPITATING •LECTRON NUMBER FLUX (TSUM), AVERAGE

ENERGY (EMN). AND ENERGY FLUX (ELF) versus UNIVERSALTIME FOR SATELLITE PASS A624

66

1010

T-Psum

106 1

ENIN10

1010 . -

ELE

106 LA157

NORTH 10:20:00 10:21:00 10:22.00 10:23:00 SOUTH

S0 km

Ki

FIGURE 4.C.4 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOP SATELLTE PASS A157

67 j6

NI

'. 1010

T-PISUM

106

10

EMN

N ~0 -

SHORIZON

A333• 106

NORTH 10:13:00 10:14:00 10:15:00 10:16:00 SOUTH210 km- E

FIGURE 4.C.5 PRECIPITATING ELECTRON NUMBER FLUX (TSUM). AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) vet UNIVERSALTIME FOR SATELLITE PASS A388

68

-' *1

1010

EMN

SUM s 1 •3

106 A232

NORTH 10:52:00 10:53:00 10:54.00 10:,55-.00 SOUTH

280 km -- W

1K - 2

FIGURE 4.C.6 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSAL

01 1

_TIME FOR SATELLITE PSA2

NOTH1052O~ 1:5:0 1:5.0 1:5:06SUT

1010

T-PSUM

106

EMN•

ELE

I ____ __A676

NCRTH 10.38'00 10.39'0,r 10:40.00 10:41:00 SOUTH

100 km -- EK 2

FIGURE 4.C.7 PRECIPITATING ,.' :CTRON NUMBER FLUX (TSUM). AVE, -%GEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A676

70

1010

T-PSUM

106

lO

1010

ELE 17j

106 _____ A751

I I !

NORTH 11:0t0 11:04:(0 11.05:00 11:06.00 SOUTH

150 km- WK - 3

FIGURE 4.C.8 PRECIPITATING ELECTRON NUMBER FLUX (TSUMI. AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A751

1-<

71

1010

i,-P I

SUM

106

'o r

EMN

0 2

,o o I

ELE -- Ai .\

.06 2I !I I•NORTH 10.27:00 10:28.00 10"29.00 10:30:00 SOUTH

0 kmK - 4

FIGURE 4.C.9 PRECIPITATING ELECTRON NUMBER FLUX (TSUM). AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A369

72

10 10

T-P h

1061 _ _

10

10 10

ELF

106 I10

NORTH; 51:00 10:52:00 10:53:00 10:54:00 SOUTH

ENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSAL

4 TIME FOR SATELLiTE PASS A657

734'

loin

T-P

SUM

10

EMN hV0 01

NORTH 11:08.00 11,09.00 11 "0.00 11 :11.00 SOUTH.,"

370 km - W

FIGURE 4.C.11 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVE.RAGE

ENERGY (EMN)o AND ENERGY FLUX (ELE) versus UNIVERSAL •

S~TIME FOR SATELLITE PASS A431

744

- - --

T°PJ - . j .- , •-

1010 1

T-P

SUM

106 __

10-

; ~FIGURE 4.C.12 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGE =ENERGY (EMN), AND MNERGY Fl UX (ELE) versus UNIVERSAL

STIME FOR SATELLITE PASS A2964 "t

775

ij

, P{

liii

T-PSUM

106

10

EMN 1

1010

ELE

1•06 ................ } JA2995

NORTH 11:54:00 11:55:00 11.56:06 11.57:00 SOUTH

530 km- Ei ' K' 7

FIGURE 4.C.13 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN), AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A2995

76

I _ _

1010

T-P A .SSUM ....

-- um

106

"•110 .- V IEMN

1010 ~-

ELE

fi;YOND1"O N A2979S~106 .

NORTH 12 10:00 12:11:00 12:12.00 12:13:00 SOUTH

300 km - E

FIGURE 4.C.14 PRECIPITATING ELECTRON NUMBER FLUX C. (,UM), AVERAGE

ENERGY (EMN), AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A2979

-- .7

Slos1

105

103

E

101

PAV

106

t, '•APAV ]-

4 I-] A518

,II I

NORTH 10:14:00 10.15:00 10:16:00 10:17:00 SOUTH

260 km- E

K -0

FIGURE 4.D.1 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (Ei).

STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A518

78

i4

?Nf~s

HIGURE 4.D.2 PRECIPITATING PROTON NUMBER FLUX (B), AVERAGE ENERGY (E),STATISTICAL ERROR iN AVERAGE ENERGY (AE'), ENERGY FLUX (PAV),

AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A641

7

,i

103

B

l1f\

10-lop

',103 1.•

101

• • :A6240

K - 1

I PAVI I I8I

NORTH 10:13:00 10:14:00___ 10:15.00 10:16:00__ SOT

iIPA km--

FIGURE 4.D.3 PRECIPITATING PROTON NUMBER FLUX '.B). AVERAGE ENERGY (E'),i

STATISTICAL ERROR IN AVERAGE ENERGY (tiE), ENERGY FLUX (PAV),

AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A624 ;

80 :

04

10" -

-El

g.I010..

. . - I

FIGURE 4.0.4 PRECIPITATING PROTON NUMBER FLUX ((B). AVERAGE ENERGY (E).

STATISTICAL ERROR IN AVERAGE ENERGY ME•). ENERGY FLUX (PAV),AND STATISTICAL ERROR IN ENERGY FLUX (A•PAV} vemus UNIVERSAL :TIME FOR SATELLITE PASS A157

1081

108

Eu

.. 1

lo,

AE1

InPAV

HORIZON . II I __ _

NORTH 10:13:00 10:14:00 10,.15:00 10:16:00 SOUTH210 km- E

K- 1

FIGURE 4.D.5 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E).STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STAIrSTICAL ERROR IN ENERGY FLUX (APAV) veris UNIVERSALTIME FOR SATELLITE PASS A388

82

i:Qo -0a-"aa

1031

- -A

E

10" "

lop

APAV

I o II•

NORTH 10:52:00 10:53:00 10:54:00 10:55:0 SOUTHa

280 km -W

K - 1

833

S~FIGURE 4.0.6 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E').

• -•-7SThTISTICAL ERROR IN AVERAGE ENERGY (AE'). ENERGY FLUX (PAV).

S~AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALS~TIME FOR SATELLITE PASS A232

83 w

10:3 i

10.

E

101

idPAV

1 9I

APAV

L ,A676

NORTH 10.38:00 10:39:00 10:40:00 1C:41.CO SOUTH

100 km - EK-?

FIGURE 4.D.7 PRECIPITATING PROTON NUMBER FLUX (B,. AVERAGE ENERGY (E).STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV). versus UNIVERSALTIME FOR SATELLITE PASS A676

84

-•-

lopl

- ----- IA °10•

i t If

A75

EN

1050 k - 1

_ _ _ _ _ K -13ARPAV _

TIM FOA7TLLT5PS A5

! !j JINORTH 11 03:00 11:06:00 11:015:00 11:06:00 SOUT.'

150 km -- W

• ~FIGURE 4.0.8 PRECIPITATING PROTON NUMBER FLUX (B). AVEP•AGE ENERGY (E),

I STATISTICAL ERROR IN AVERAGE ENERGY (AE'), ENIERGY FLUX (PAV).

AND STATISTICAL ERROR IN ENERGY FLUX (&PAV) versus UNIVERSALTIME FOR SATELLITE PASS A751

iVa-

loý

103

lop

105 . ..

PAV

106

NORTH 10:.27:00 10:28:00 10:29:00 10:30:00 SOUTH0 km

FIGURE 4.D.9 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E).

STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV) verss UNIVERSAL

1086TIM FR ATELIE ASSA39

____I

10:30:0 SOUT

IdIW

10

H ON

1 106

f" H -1:0 10520010530 10:40 OT

APAV~ -' 5

FIGURE 4.D.10 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY ( E).STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAy).AND STATISTICAL ERROR IN ENERGY FLUX (.APAV) versus UNIVERSALTIME FOR SATELLITE PASS A.657

87

lop

los _ _ _ _ _ _ _ _

01

AN TTSTCLERO NEERYFU VAAV veu UNIVRA RoP______________TIME____ FOR SAAGNETTC ASPEC A431

NORH 1:8:0 1:9:Q 1:1:0 1:1:0 SUT

FIGURE 4.0.12 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY OF).

STATISTICAL ERROR IN AVERAGE ENERGY (CE), ENERGY FLUX (PAV),

AND STATISTICAL ERROR• IN ENERGY FLUX (APAV) vresus UNIVER•SALTIME FOR SATELLITE PASS A2964

89 •

;4

108 ,'..,

'A. IC:^ • ' '

- . I%

1013..

AE

106 "

PAV

APAV "

01 ~ A2995

NORTH 11:54:00 11:55:00 11:56:00 11:57:00 SOUTH

530 km - E

FIGURE 4.0.13 PRECIPITATING PROTON NLAMBER FLUX (B). AVERAGE ENERGY (E),STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A2995

90

* o's .. . . l .. -

103

IL

I 1lol HOIZN

10 6

APAV~ ~HORIZ ON i AM 9'7o I I

NORTH 12:10:00 12:11:00 12:12:00 12:13:00 SOUTH

-W km- EKt-7

FIGURE 4.D.14 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E),STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV),AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A2979

91

7; 1 _

. 100

< ->

. 10

U)

BELOW 3-1106 le 108 091010THRESHOLD TSUM -- i-2 S71 ster'.

LU1a I I

100U)

z -

wa - - ,;N.Z 10U)

0II

BELOW 31100 10e 108 109 1010

+THRESHOLD ELE - keV/cm2 s ster

.~J1 0 0 IRADAR> l "-AURORA PRESENT

I.- NO RADAR; 10 AURORA PRESENT

C" A518

<*~*'K - 0

1 -~SimnultareitV 7 minutes

0 5 10

EMN - keV

FIGURE 4.E.1 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITAT!NGEL.ECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A518

92

. 100

> g

z S10"J •~.j .

BELOW 3x10 6 10e 108 109 1010

THRESHOLD TSUM - cm- 2 s-1 ster-1

j 100

"C

o I I.BELOW 3x108 lO 108 109 O101

THRESHOLD ELE -- keVlcm2 s ster

,100 I-RADAR

• -- _AURORA PRESENTw -iI- -- -- NO RADAR

-_ 10 • AURORA PRESENT"U)

0A541

THEHL 5 L 10/c S2taet -1 minute

EMN - keV

FIGURE 4.E.2 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATINGELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A64I

- - _ ~ -4 10

-j 100

cc

z. 10

In

0

BELOW 3x106 107 108 109 1010

THRESHOLD TSUM - cn- 2 s 1 Ste.-i

100

wI-.

z

0

BELOW 3x10 6 107 108 1010

THRESHOLD ELE - keVlcr 2 s $teR

j100 I_j 10 __-- 8-"'7 RADAR

:> F-1 AURORA PRESENT

t- •• IL•-•r,•'• NO RADAR

10 "AURORA PRESENTIn

K -•" ,•,• . •Simultaneity -- Good

0 10

EMN - keV

FIGURE 4.E.3 HISTOGRAMS OF LOGARITHMIC OCCURRENCE veasus PRECIPITATINGELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A624

94

-; --

AI

Z 100

10

SBELOW 3.106 1oe 108 log 1010THRESHOLD TU-& 1 s&

:>cc

:100

z< 10

U)

w

BELOW 3.106 107 108 log 1010

SELE - keVkcm 2 s ster

_j 100 •_.•< -- 0 RADAR

> -- AURORA PRESENT-- -NO RADAR 75

- 10 AURORA PRESENTU)

~ ,~-,A157

U) K ='2A]4 ~., .' Simultaneity - Good

010 5 10

EMN - keY

FIGURE 4.E.4 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATING J;

ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A157

95

0 0 ~~~w -A;

-'S

-J 1)00-

z

,,g,

'- -- -

10..,

0 . ..... "" k ........*'•':

BELOW 3.10 6 1(7 1o8 19 10 10

THRESHOLD TEUM - c-cm2 s-1 stef1

"J100

AURORAPRESEN

I- -- NO RADARz -- • AURORA PRESENT

.ir

10 ~~

4 1 •;•• -••, ••, - Simultaneity -- Good

00 5 10

EMN - keY

FIGURE 4.E.5 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATING

ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A388

96

> F- UOA RSN

- -u

100

ILl

0BELOW 3x10 6 107 108 109 1010

THRESHOLD TSUM - cm-2 s 1 ster"1

-j 1004>

z10

CL)

BELOW 3x 06 107 108 log i010THRESHOLD -ELE -- keV/cm2 $ ster

S100 R A D A> - F AUROR•A PRESENT

W-•10 • • AURORA PRESENI[

CL)THRE-HOL ELE-NOy/RADARte

-K

rn Simultaneity - Good

S~0 00 5 10

EMN - keV

FIGURE 4.E.6 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATINGELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A232 N

97

1001

> -

Z 0

0-

BELOW 3.10 6 10 i8 109 101THRESHOLD TU m 2 slse-

I100 RA R

>JARR RSN

NORAA

Siutaet -Go

FIUE10UOA RSNOCURNE essPECPTTN

ELCRNNME LU TU) NRY LX(L) NAVRG.NRY(M)FO AELT ASA7

U98

,j 100 -

10lc--

w~ 10

BELOW 30106 10e 108 109 1010

THRESHOLD TSUM -cm-2 7-1 ster-1

100A

• z

4 0

0 °

BELOW 3x`10 107 108 10 101

THRESHOLD ELS - keVlcm2 s ster

100

> F1AURORA PRESENTcc

4/

-- - |•'P0 NO RADAR. Z10 • • AURORA PRESENT

to

K -

~ 0

0 5 10A

EMN - keY

FIGURE 4.E.8 HISTrOG3RAMS OF LOGARITAMIC OCCURRENCE versus PRECIPITATING

ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PAS A751

99

wýj

.j100

>

cc

z -I I q•! I ,

BELOW 3x106 107 108 log 1010

THRESHOLD TSUM -- cm-2 s-1 ster-1

I-2

.100

- -o<4 I l I

>ui

-- .-n.*___BELOW 3x10 6 107 108 109 101

THRESHOLD " 4 : cO)tC ~ j RSNll

0IBELOW 3x10 6 10 181911ELE - keV/cm s ster

J100S1I-RADAR

S1 AURORA PRESENT

S1o • •:• • NO RADAR10 , . AURORA PRESENT

"CL Y •y.. A369

z Simulitaneity-i minute

010 5a-L 10EMN - keV


ELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A369

100

100

-I100

w

I-z10

w

0.

BELOW 3.106 107 108 109 1010THRESHOLD TSUM - cm"2 s- 1 ster-

-J100

S10

BELOW 3-10 107 108 109 100TELE -- keVcm2 s ster

z <'1°°" _- [ RADAR: > __AURORA PRESENT 2

U)

I- "- -- NO RADAR

10 A P

ELE K -k592

300

FIGURE 4.E.10 HISTOGRAMS OF LOGARITHMIC OCCURRENCE veADAs PRECIP;TATING

ELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), AND

AVERAGE ENERGY (EMN) FOR SATELLITE PASS A657

Szoz

101i

00 ~Samuta~ety -Goa

• - , I i'• " •;.. .+ +"+ '" • " ',;+ '+' •:'+ +-".• •+ .+m .- • +:.-•: +-.,,•..+ ... ., .. + .. +++.+.+ • +.10+.

j 1004c

7- 1Ion

0

BELOW 3x10 10 i8 109 100THRESHOLD TSUM -cmn

2 S71 ster-1

100

zZ 10-

0-

BELOW 3-io 1 _C I07181911THRESHOLD ELE keVlcm2 s ster

100 RADAR

NO RADARz- 10 AURORA PRESENT

C--"d ~A431

In~ = K5'14M1 Simultaneity -3 minutes

00 510

EMN -keV

FIGURE 4.E.11 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATINGELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A431

102

S100>

z '

0

BELOW 3.106 10 log10 1010THRESHOLD TSUM -fj- S7 ster-I

4100

1 0

BELOW 3-106 10 109 10100THRESHOLD ELE -- keVlcm2 s ster

.j100 RADAR<~ AURORA PRESENT

Lu~ NO RADAR10 AURORA PRESENT

A2964~' K - 6

~, ~Simultaneity -Fair

00 5 10

EMN - keV

FIGURE 4.E.12 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIP!TATINGELECTRON NUMBER FLUX (iSUM), ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN;t FOR SATELLITE PASS A2964

103

j 100a>

::z ~ 100

LU

C

I-.

IBELOW 3.c-§ 10e 108 109 1010

THRESHOLD TSUM - cm -2 s -1 ster -1

.j100 -

-10

u,

-F

BELOW 3-106 107 108 lop 1010

THRESHOLD ELE - kecm2 s' ft- r

.<100 _ -I- • RADAR

•> AURORA PRESENT

Sus

uJ I • •-- • NO RADAR-- 1zr-' AURORA PRESENT

S1io iutoet •• oCL

10toK=

0 5 10EMN - L kcV


ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A29

100

tr 11-10tr

0.-

•N -, R.

BELOW 3x106 107 0 109 100

THRESHOLD TSUM -- cm-2 s-1 ster-1w 100

0-0. - -- - I-

BELOW 3x10 6 107 108 10 1010

THRESHOLD ELE - cm 2 st

< -- -- RADAR> - AURORA PRESENT

z -- ~ NO RADAR

S•.,,••:Sinultaneity -- Gnod

0 5 10

ELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A2979 _

105

....... ......

5. COMPARISONS OF RADAR AURORA

WITH MAGNETOMETER AND ALL-SKY-CAMERA DATA

In this section two other kinds of data correlations that may shed

light on the understanding of radar aurora are presented. The first

study is a comparison of magnetometer data with radar aurora. The second

comparison is between radar aurora and visual aurora as recorded with

all-sky cameras.

5.1 Radar-Aurora Magnetometer Comparison

On 24 March 1969 radar data were collected for approximately

5 hours for the purpose of seeking a correlation between radar echoes

and magnetometer response for a magnetometer placed below the region from

which radar echoes were being obtained. The radar antenna was directed

toward the ionosphere at an qltitude of 110 kilometers on the magnetic

field line that intersected a magnetometer operated by the University of

Alaska located at College, Alaska. We present here a composite figure of

radar data and magnetometer data; the composite is given in Figure 5.1.

The upper portion of the figure gives magnetometer response versus Green-

wich Meridian Time for the Z and H compinents of magnetic field, At the!t

bottom of the field are three amplitude-versus-time traces. The ampli-

tudes are given in dBm. These are radar signal amplitudes obtained witb

the 139-MHz radar. The amplitudes were measured using a 40-km range gate.

The rangc gates were centered on the magnetic field line through the

magnetometer (center trace) and at ranges less than and beyond the mag-

netometer field line--as indicated in the figure.

Visual inspection of the data presented in Figure 5.1 does not

show an obvious correlation between behavior of these two instruments

porECEWNS PACE BAK 107I

I____

(radar and magnetometer). The magnetometer samples ionospheric currents

over a fairly large north/south extent of the ionosphere. It is probable

that our three radar range gates are so close together that the magnetom-

eter averages over effects in all three gates. It must also be remembered

that magnetometers respond to current in the ionosphere; it is believed

that conditions for the structuring of ionization that leads to radar

scattering requires ionospheric electric fields, not electric currents.

Therefore lack of an obvious, detailed correlation may not be surprising.

5.2 Radar Aurora--All-Sky-Camera Comparison

The question of relationships between visual aurora and radar

aurora has not been unambiguously resolved. There is a tendency, we

* believe, for radar aurora to be located near to but not necessarily coin-

cident with visual aurora. This relationship has been found on occasion

during periods of fairly stable, quiescent aurora. The study of the re-

lation of radar aurora to visual aurora currently requires stable, well

defined visual arcs and stable radar aurora. This is because it takes

several minutes to accomplish radar scans to map radar aurora. Many

aurora will move very great distances during this period of time. It is

expected that with the addition of the phased-array radar to the Homer

facility, more dynamic aurora can be investigated to seek visual radar

relations.

Our experience has shown that radar/visual correlations are

only meaningful if the visual data are obtained by triangulation between

at least two camera sites. For Alaska this requires inordinately good

weather. Otherwise, cloud cover prevents successful acquisition of all-

sky data.

Figures 5.2 and 5.3 provide two examples of radar visual corre-

lations. These data were obtained with the 139-Mfz radar. The data for

108

Figure 5.2 were obtained at 12:19 GMT on 24 September 1969. The data for

Figure 5.3 were obtained at 12:28 GMT on 10 October 1969. In Figure 5.2,-J$

two visual arcs are indicated; these were located by triangulation between

all-sky photographs taken at Ft. Yukon and at College, Alaska. The re-

gions of radar return are indicated by the bars located generally to the

north of the northern visual arc. We note that the radar returns seem

to be adjacent to, but not superposed upon, the visual arc. We note that

in the eastern end of the northern visual arc, the visual arc seems to

bifurcate the radar echo. The data of Figure 5.3 generally show radar

returns that are also bifurcated by the visual arc. '7

It has been speculated2 that since radar aurora is thought to

be produced by high electric fields, the high conductivity produced by

ionization generated by the precipitating particles that cause visual

arcs will short circuit these electric fields. In the absence of the

electric field, plasma instabilities cannot grow to produce structured

ionization that is needed for radar scattering. These visual/radar com-

parisons seem consistent with this speculation.

109

109

z

LU

0I- LU

Z w L0- I

E EXu X00

w 0 0 UCD 1

4 - - I

UJO

j2LL

co V

CD 9L

N 0 ,C

110

I It

4-r

cc

I.-

U-

-~ 1 11

Cl!-

IE

00

i~t '1D

ixrl

112

6. STUDY OF SATELLITE DATA

It is now universally agreed that radar aurora is produced by scat-

tering of radar energy from field-aligned irregularities in E-region

ionization. There are several important questions that are continually

being addressed. Some of these are as follows:

(1) What produces the E-region ionization (during nighttime)?

(2) What mechanisms produce irregularities in the E-region

ionization?

(3) What is the detailed relationship between radar aurora

and other auroral phenomena?

In this section we study some of the physics and chemistry of pre-

cipitating particle interactions in the ionosphere as background to in-

terpreting the satellite data. We then concern ourselves with preliminary

interpretation of the data presented in Sections 4 and 5 as a start in

clarifying our understanding of the problems posed as questions above.

6.1 E-Region Ionization Produced by Precipitating Electrons

A question of concern in radar aurora is whether precipitating

particles produce the ionization that is necessary for radar clutter

echoes when the E-region is not sunlit. Daytime (sunlit) radar aurora

can be produced by structuring of normal daytime E-region electrons.

We have pursued this question in relation to the satellite precipitation

data in order to better understand the production of nighttime E-region

ionization required to produce auroral radar returns. The data provided

to rs by Lockheed only gave the average energy of precipitating electrons.

113

The detailed spectrum based on their five energy-selecting channeltron

counters was not made available to us. Figure 6.1 presents dataP giving

the altitude of maximum ionization rate due to precipitating electrons.

The data in this figure are for precipitation at a geomagnetic latitude

of 670. The pitch-angle distribution is assumed to be isotropic over the

downward hemisphere. The data of Figure 6.1 are for a monoenergetic

electron energy spectrum.

* The precipitating electron spectra are not monoenergetic. The

proton data to be presented later, as well as the rate of photon-production

data for an exponential electron spectrum,7 reveal deeper penetration

than indicated in Figure 6.1 if the altitude of the effect is referenced

to average particle energy. That is to say, had we used an exponential

spectrum for the electrons and plotted ionization rate versus average

electron energy, then the curves in Figure 6.1 would be moved to a lov.er

altitude.

Figure 6.1 indicates by cross-hatching the band of altitudes

over which radar aurora is most frequently observed. For monoenergetic

electrons with isotropic pitch-angle distribution, energies must exceed

5 keV before ionization is produced in the altitude region of prime

concern to radar clutter. Because electron spectra probably have a spread

in energy when the EMN from the channeltron spectrometer exceeds 2 kV

or so, we presume that ionization will be produced by the electrons in

the altitude region of the most frequent and most intense radar clutter.

Since the satellite data provide an absolute flux, it is pos-

sible to estimate the E-region electron density produced by the precipi-

tating particles. Figure 6.2 provides some results of these computations.

Altitude in kilometers is plotted versus ionospheric electron density for

various precipitating electron fluxes. These electron densities are

computed at each altitude using a monoenergetic electron beam that pro-

duces its maximum ionization rate at that altitude. Thus the computation

114

J-

at 100 km altitude assumes a monoenergetic electron flux of 17 keV %see

Figure 6.1). The calculation for an altitude of 130 km., for example,

assumes an electron energy of 2.5 keV.

The steady-state electron density in these calculations was

computed assuming that unperturbed atmospheric chemistry is pertinent.

The electron clean-up mechanism that dominates the chemistry is disas-

sociative recombination. This reaction varies with altitude, according

to data summarized elsewhere. 8 The reaction is also temperature-sensitive

so that if the auroral precipitation or the auroral electrojet heat the

ionosphere above normal temperatures, higher electron densities than we

compute may be produced.

The radar clutter data do not directly reveal the electron

density in the scattering regions. Under very special circumstances one

can deduce approximate values for this quantity (see Ref. 2). When this

is done, values of nighttime E-zegion electron densities between 10 and6 3

106 el/cm are found. To produce these densities, accorcing to Figure7 2

6.2, fluxes equal to or greater than 10 el/cm -s-ster are required. The

electron counter threshold, below which Lockheed advised us not to use6

the data, is 5 X 10 . Figure 6.2 indicates, therefore, that when electron

precipitation is measured by the satellite above its detection threshold,

the precipitation is nearly capable of producing the electron densities

that seem to be needed.

Another important question is that of persistence of the ioni-

zation produced by precipitation. If the precipitation suddenly turns

on, how long does it take until the ionospheric electron density reaches

the steady-state electron density as determined from Figure 6.2? Figure

6.3 presents altitude versus chemical time-constant for the electron-

precipitation fluxes described by Figure 6.2. In the altitude region of

concern (105 to 115 kilometers), Figure 6.3 indicates that chemistry

115

lifetimes are shorter than 20 to 30 seconds for the satellite electron

precipitation threshold of 5 X 106. We conclude from this figure and

Figure 6.2 that the electron densities produced by electron precipitation

as measured by the Lockheed satellite are great enough and can be pro-

duc~d in a sufficiently short time to explain the presence of electrons

at E-region altitudes during the nighttime sten auroral clutter is ob-

served. We study Later whether there is spatial overlap of ineasureable

electron precipitation and radar aurora.

6.2 E-Region Ionization Produced by Precipitating Proton

As with electron precipitation we ask the question, do precipi-

tating protons produce enough E-region ionization--in the absence of

sunlight--to explain radar auroral clutter when clutter and proton pre-

cipitation spatially overlap? To investigate this we have plotted data

obtained from Chamberlain. 9 The data are for an exponential (in energy)

proton spectrum, and an isotropiv pitch-angle distribution in the downward-

going hemisphere. The results are presented in Figure 6.4. Also indi-

cated is the altitude at which the ionization production rate is 10 percent

o0 the peak value. The upper-altitude 10-percent region is so high it

was not plotted.

To aid in ,mderstanding penetration of a continuous spectrum

compared with the monoencrgetic spectrum that was used for tthe electron

£ precipitation, the stopping altitude for monoenergetic protons with zero

pitch angle was also plotted in Figure 6.4. The isotropic pitch angle

distribution would tend to raise the altitude of peak ionization with

respect tc, this lntter, zero-pitch-angle curve. The high-energy tail of

an exponential spectrum moves the altitude of peak ionization rate do*rn-

ward into the atmosphere, if one plots peak ionization rate versus average

proton energy. We note that the proton average energy must exceed about

116

10 keV in order to produce ionization in the altitude band of most fre-

quent and intense auroral radar clutter.

As we interpret the data provided in Ref. 9 we can only compute

the average electron density produced by a proton exponential energy

spectrum with average energy at 32 keV. Using the reaction rates of Ref.

8 we have plotted the electron density versus proton flux in Figure 6.5.

We assume that a background ionization density of 105 el/cm3 is needed

to explain auroral clutter; Figure 6.5 indicates that the proton flux

vould have to exceed 5 X 10 in order to produce the needed ionization in

the absence of solar (daytime) E-region ionization.

It is important to note that precipitating protons can undergo

a process that is not available to the electrons. A precipitating proton

can pick up an electron from an atmospheric atom and thereby become a

neutral particle. Once the proton is neutralized it is no longer con-

strained to follow the earth's magnetic field lines. On deeper penetra-

tion of the atmosphere, the neutral proton may again become ionized.

By this process the protons can migrate across magnetic field lines.

Therefore, proton fluxes that are measured by the satellite at altitudes

of about 400 km may be spatially spread by charge exchange by the time

the proton reaches altitudes of 100 to 120 1m. As a result, a localized

peak in proton flux measured at an altitude of 400 km may become spatially

spread by 100 km or more by the time the protons reach E-region altitudes.

Estimates of E-region ionization that may be produced by proton precipi-

tation as measured by the satellite must be spatially averaged along the

satellite path in some manner. We have not pursued this question in any

iaore than qualitative depth at this time.

117

4

41

6.3 Correlation of Radar Echoes with Satellite-Borne Particle-

Precipitation Measurements

In this subsection we compare some results of the simultaneous

measurement of radar aurora and particle precipitation. In particular,

we compare locations of peak energy flux with locations of radar aurora.

In some sense, this kind of measurement may be redundant with radar/all-

sky camera comparisons, in that it would be expected that spatial peaks

in satellite-measured electron precipitation would give rise to discrete

auroral forms that would be recorded by all-sky cameras as visual arcs.

Other investigators have not been able to well substantiate this relation,

but few believe the relationship should not exist. We shall also study

the question of whether sstellite-measured precipitation is sufficient to

provide the ionization that must exist in regions from which we obtain

radar echoes if these echoes are to be observed.

6.3.1 Correlation of Regions of Peak Energy Flux

with Locations Radar Aurora

The data from 14 satellite passes have been studied to

seek a variety of correlations. Table 6.1 presents one set of correla-

tions of these data. The first column of Table 6.1 identifies satellite

pass number.

We have categorized the type of auroral event that was

occurring in Column 2 according to the nature of the satellite precipi-

tation data. If there is no symbol, the event was a weak precipitation

event, meaning that the particle fluxes were quite low but were measurable

in some parts of the trajectory. If, in our judgement, the energy flux

of electrons (not the number flux) was large, then we have placed an E

in this cclumn. if proton precipitation was weakly observable over a

reasonable portion of the orbit, "we have placed wp, meaning moderate to

weak total proton euergy. Three events have been labeled with a P because

118

4.4

0.a a

a. .. 0 54

.0 0) VC 0 a c .0

~.va S. S o SE.i 10 4i :1V 0 0' S

4S. 00 k 0 0000 '00 0 v 130 * 40 000 0;a0 0 0 0 0 0 0 a a 0 00 c 0

a W 0 C0 0 ~S. 0 ~ol 0 fu 00 00- 4.

V V0 0C a a S. -.0 a i 136

0 ra 4. S. S.0 I 4. X 0 Ixix aCJO0 C3 34 w a

3. C. - - 4 - - - - 0 0

00

'0 S. a4 t..0 X.

.0

0 ~0 00 I-

93 0i CS. S

03 z. C-3 N 00 03 v V l r04.

0~. =~l a~.S aa aS. 0

0a ~0

0 .0 c~0 .0 O- ~ .40 0 a C3 4. 4

4.4 N N O 0S. .-4 0 S. S. 0 ~ 4

rj 4 0 M4.S Sc-: C . 1 0 -ý ' '0

O.- 00 0 04 - o Vi> - = rE- . IV. S4: 3 a C3

V' 0 01 -q L) 0 S. >4 4 IQ m2x : 4a 0. 0- ~ 0 '

4- aý13 -j

0 0 a a3

C:

0 C-VI 3 r

U

+s. ca03 0 - - -N N C C OC C- M~ 0-

0 '0 41

0 0

m V4.V t l Q s

z

u S.

0 - w mCaCIS v

119

25

the proton energy flux was very large. For our own work we have named

these "proton events."

The next column gives the College, Alaska three-hour

magnetic index for the period of the satellite pass.

In the electron-precipitation columns we have defined

collocation of electron energy peaks and radar aurora as coincidences.

This means that radar aurora seemed to be located in the same regions in

which there is a local maximum in the energy flux of precipitating elec-

trons. We have also counted the number of isolated radar aurora and peaks

in electron energy flux that were not associated with each other, and have

called these anti-coincidences. We have done the same for proton pre-

cipitation. The last column gives a measure of the simultaneity of the

radar and satellite measurements (on some passes the ephemeris data

available at the time that our data were reduced were less accurate than

we would have desired).

The determination of these correlations is extremely

subjective. In some areas there will be many, closely spaced, intense

precipitation peaks. In some cases the radar auroral boundary may be

located at the particle-precipitation peak. We have tried to be consis-

tent in calling these coincidences. The subjectivity of the interpretation

srsuld be evident by the way ue have filled out our columns.

Our attempts to redo the electron precipitation columns

always produce a slightly differing coincidence set--that is to say, our

estimation procedure is very subjective. This is very disturbing, but

we have not discovered a way to uniquely define precipitation peaks. Our

reevaluation of coincidences--though very subjective--does show a tendency

for radar aurora to be anti-coincident with peaks in electron-energy

precipitation except for the cases of very active aurora.

120

Examination of Table 6.1 as given here reveals the

following:

* In the auroras that we have not categorized as

proton events and not labeled as P, there are no

coincidences between radar returns and peaks in

proton precipitation.

* There are some coincidences between peaks in

electron-energy fluxes, but there are many more

anti-coincidences.

. For two of the proton events, since radar aurora

I is nearly everywhere, there are many peaks in

energy deposition, so that coincidences are

everywhere and therefore may not be meaningful.

* The electron-coincidence enumeration suggests to

us that coincidences between peaks of electron

precipitation with radar aurora is almost a random

phenomenon. There appears to be no clear-cut,

absolute trend, though there are more "valid"

anti-coincidences than valid coincidences.

In summary, we believe the data in Table 6.1 indicate

that except for intense proton events, radar aurora is not coincident

with peaks in proton energy flux. Peaks in electron precipitation may

or may not be coincident with radar aurora; we have not discerned a

meaningful trend, although there seems to be a tendency toward anti-

correlation.

We believe ionospheric electric fields play a role in

causing radar aurora. Electric fields are caused by convecting magnetic

fields. If the ionospheric ionization is great enough, then magnetic-

field convection is slowed in the highly conductive ionosphere. Nearby,

121

where ionospheric electron densities are lower, large electric fields

are not short-circuited by high conductivity. One might therefore expect

that at locations where particles are heavily precipitating, ionospheric

electric fields will be short-circuited and there will be no structuringof ionization into irregularities and therefore no radar scattering.

Thus, during moderate auroral activity, regions of most intense (by par-

ticle number) precipitation should not coincide with regions oi radar

echoes.

On the other hand, if magnetic-field convection becomes

extremely intense, even in regions of large precipitation (and therefore

high E-region electron densities) electric fields may reach and exceed

a threshold for production of plasma instabilities that lead to generation

of E-region irregularities.

Therefore correlation between peaks in precipitation

intensity along the satellite path and no radar aurora at the same loca-

tion might be found at times of moderate or weak auroral activity. During

strong aurora one might seek and find an inverted relationship. Table

6.1 does not support this hypothesis.

As an alternative approach to organizing data correspond-

ing to the above intuitive guidelines, the histograms of Figures 4.E.1

through 4.E.14 were made. If the above ideas concerning electric-field

4 short-circuiting in regions of high energy precipitation were valid, then

the histograms should show more energy precipitated (larger ELE values)

in regions where no radar aurora was observed. Examination of these

histograms is consistent with this hypothesis except for five passes.

Passes A518 and A641 seem inconsistent, but the time simultaneity was poor.

Passes A232, A751, and A369 are not consistent with the hypothesis.

Passes A2995 and A2979 are consistent in the sense that these passes

occurred during very active aurora.

i122

I A

i- Examination of the other histograms of Figures 4.E.1

through 4.E.14 show no other obvious consistency or trend.

An additional attempt to reveal some relationship between

precipitating electrons and the peesence or absence of radar aurora in

S~the same region of space was made by computing average quantities from

the histograms of Figures 4.E.1 through 4.E.14. Figures 6.6 and 6.7

provide summaries of averages obtained from the histograms.

The top subfigure of Figure 6.6 presents average TSUM

for each satellite pass plotted against sequence number according to

ascending three-hour magnetic K index. Open circles are the average

TSUI for the regions of the satellite pass where radar echoes were simul-

i taneously observed; solid points are for the regions of the pass whereII no radar echoes were observed. On Pass A676, which is number 7 in our

* ordered sequences of passes, no electrons were measured above counter

t threshold in regions where radar echoes were observed. Thus, no point

Sis plctted there for radar echoes. The middle subfigure presents the

same kind of data for the average electron power (ELE). The bottom sub-

figure presents the average electron energy as a function of satellite

pass according to ascending three-hour K index. In the bottom figure we

have provided a linear regression fit to the data points. The fit for

the data without radar echoes is shown as a solid line. The fit to the

points corresponding to radar echoes present is shown as a dashed line.IA

The fit was made with sequence number as abscissa; the fit serves only 3

to dramatize similarities and differences between average electron energyInin regions of radar aurora and in regions of no radar aurora.

Fig-re 6.7 plots the ratio of the various average quan-

tities of Figure 6.6 versus satellite sequence number. The top subfigure

in Figure 6.7 presents the ratic of TSUM from regions of radar aurora to

the average TSUM in regions of no radar aurora. This ratio is then

plotted versus satellite sequence lumber. The middle subfigure presents

123

this ratio for ELE. The bottom subfigure presents the ratio of averageI precipitating electron energy in radar regions to that in other precipi-

tation regions.

The top subfigure of Figure 6.7 gives the impression

that, on the average, the precipitating electron number flux is lower in

regions of radar echoes than in other regions. In fact, this i3 only

true in 8 of the 14 passes. However, the geometric mean of the 13 useabl.e

passes (Pass A676 had no electron precipitation in regions of radar

aurora) is 0.81, meaning that in expectation the average number !lux of

the precipitating electrons in radar scattering regions is 80&r of that

for electrons precipitating elsewhere.

The data from Figures 6.6 and 6.7 cencerning the energy

flux (ELE) seem to show more definitive differences between precipitation

characteristics where radar echoes wo're obtained and where they were not.

During only three passes of 14 was ELE greater in regions of radar aurora.

If this ratio was a random quantity, then the probability that three or

fewer passes as we have found would have ratios greater than 1.0 is only

2.9%. The ratio measured on two of these passes was only slightly greater

than 1.0. The geometric mean of the 13 calculable ratios is 0.55, sug-

gesting that, on the average, the energy flux is only half as great in

regioDs of radar ehcoes as it is elsewhere where precipitation inxensity

is above the satellite threshold. This averaged characteristic is con-

sistent with our often-stated hypothesis that more precipitation produces

i: .. e ioni. )n, resulting in short-circuiting of ionospheric electric

fields.

The data plotted in the bottom subfigure of Figure 6.6

show a decided precipitating electron energy trend with 3-hour magnetic

K index. The data suggest--and the corresponding subfigure of Figure

6.7 further support--an indication that on the average, electrons that

124_ _

; I _ _ _

precipitate into regions from which radar echoes arise have lower energies

. than do electrons that precipitate elsewhere. On 8 of the 13 useable

passes the average precipitating electron energy was lower in regions of

radar aurora than elsewhere. On tuo passes the average energies were the

same 4n the radar and non-radar regions. On three passes the average

Selectron energy was greater in regions of radar aurora--but the largest

ratio computed was only 1.08--which value just barely exceeded 1.0. From

these data we compute a geometrical mean energy ratio of 0.84.

6.3.2 Production of Nighttime E-Region Ionization

by Particle Precipitation

The previous section showed that regions of peak energy

deposition did not seem to be located, necessarily, in regions where

radar echoes were obtained. In this subsection we study whether any pre-

cipitation at all was occurring in the radar-reflecting regions, to de-

termine whether particle precipitation can explain the existence of

nighttime E-region ionization necessary for these reflections.

All 14 satellite passes occurred during E-region darkness.

Radar aurora requires the presence of E-region ionization. Studies of

radar/visual aurora correlation show that the visual forms do not neces-

sarily correlate well in position with radar raturns. One is therefore

led to wonder what produces the iopization that is necessary for radar

scattering. The concern here is whether there is any particle precipita-

tion at all in regions from which auroral echoes are obtained.

The particle-flux precipitation in regions from which

radar echov- -ere being obtained h.. *een investigated even if these

were not regions of peak energy flux. Table 6.2 presents these results.

On two of the passes listed in the table there were radar echoes from

regions in which particle-precipitation levels were below the detection

j threshold of the satellite. On pass A641 the simultaneity of the

125

I_- ~ ~ .- --- 3Z

S.~0 0 4J t a4J ac r= 0,.

E- -44 4'-4 &0. 0 M 0sc ý VS1JS JS ' 0= 0 4 4 0 0r

.0

C6 0 U00

CS 4 0 0 0 0 00 0 0 0 0

0

S..0

0 a 0

0 a m m w

0, 0 0 0 0 0 0 0 0 0 0 0 0 0

U 0

v

~~~ . 0 0 C02 N' fa to -40 - ) ~ - '

cz E oEa W .1 4 o t0m4

4 '

0

012

measurements was poor due to poor pre-pass ephemeris data, so that com-

parisons probably are not too meaningful. On Pass A388, 11 of 12 satel-

lite data points showed no detectable electron or proton precipitation

above our imposed thresholds where radar echoes were being obtained.

However, in this case the particle flux nearby was decaying very slowly

as the satellite approached the scattering regions. We are led to believe,

but cannot prove, that the parLicle flux was below, but very near, the

threshold detection level. On Pass A2979 the satellite data supplied to

us by Lockheed does not extend as far as our rndar echoes. Thus, lack

of displayed precipitation in the most southerly portion of the radar

echo is not of any significauce.

We estimate that, provided particle energy is great

enough to penetrate to 110-km altitude, the threshold detection flux of

the satellite counters produces at this altitude an electron density of

about 7 X 104 if the particles are electrons, and 2 X 104 if the particles

are protons. These data suggest that there was some particle precipita-

tion flux wherever radar aurora was found. The data of Table 6.2 and

our othe- inferences imply that the nighttime, widespread particle pre-

cipitation as measured by the OVl-18 during our 14 passes is adequate to

produce the ionospheric electron densities needed to explain observed

radar aurora.

6.4 Discussion of Correlation Study and Other Possible Relationships

When comparing particle precipitation with radar echoes, the

question arises whether one should compare energy flux or particle number

flux deposited. We presume but cannot prove that the energy flux is a

measure of the local strength of the auroral disturbance. From a radar

point of view the particle-number flux is a measure of the generation of

E-region ionization.

127

The particle energy spectrum is an indication of the depth of

penetration of the particles; thus it is an indication of whether parti-

cles can penetrate deep enough to produce the ionization at altitudes

C where radar aurora is observed. Furthermore, the deeper the penetration,

the lower the altitude where ionization is produced, and the greater

the Hall conductivity and the normal magnetic-field-independent conduc-

I ti?ity. As a result, the total, local ionospheric conductivity becomes

much greater, so that magnetic field convection in these regions will be

suppressed.

Conductivity is greater for large particle fluxes coupled with

deeper penetration; this may mean that high conductivity is nearly cor-

related with higher energy fluxes. We suspect, then, that the energy-

flux peaks in the satellite-pass profiles would correlate best with

magnetometer data that record location and intensity of the auroral

electrojet.

Past correlation work between location of visual and radar

echoes 2 has shown that radar echoes are often adjacent to but not super-

posed upon visual arcs. This suggests that electric fields are lower

inside the arcs because the high conductivity of the arc short-circuits

the E-field. Barium-cloud releases1 ° as well as the satellite-measured

electric field 1 1 also suggest higher electric fields near to but not in-

side visual vrcs. Thus, a correlation between energy flux and radar

echoes is a proper correlation to study. Our results indicate no dis-

cernible trend, though there may be a slight tendency to be anti-coincident.

The data of Figures 6.6 and 6.7 show that on the average the precipitation-

energy flux is lower in regions that give rise to radar echoes than in

other precipitation regions.

A radar/particle-flux precipitation correlation is meaningful

in that we need a source for the ionospheric electron density. Our

128

I__

work has indicated that widespread particle precipitation as measured by

the Lockheed satellite is adequate to produce the needed ionospheric

electron densities.

Some aspects of the data presented in this report have not been -5

studied in depth. It would be important, we believe., to digitize the

radar auroral echoes and compare absolute echo amplitude with precipita-

tion characteristics. Our estimates of particle-stopping altitudes and

resulting ionospheric electron densities, derived from work by Rees 6' 7

and Chamberlain, 9 need to be altered to agree with spectral distributions

that are appropriate for the actual satellite-measured precipitation.

Data that were recorded by Lockheed but not uade available to us concern

the shape of the electron spectrum. That spectrum is sampled on-board

the satellite at five energies. The data made available to us concerned

only the absolute flux and the mean energy. Energy-distribution details

would be of considerable value.

1N

129

~~Lj

cc 7 )

~0 - 0

25. / A1 ' 00

-J -

22 0.

22W 0 !

EC. -0.L

. 0~

LW

0 00o

cc

00

0Z l - 0

L4% 0

x Z ILo0 0 =0-L

< W

¶3 N 0 WZD alt z 130

o- LU

LL

* z0

o ~ir

LLcci

8 2 LU

- - -0W~g-31RLI~¶~z

Ul 0

Ut-U

U. 0

EZ

~UJ0

IL0

w E-

0 L

~C

z ~oI- %-~~. 0)4- 0 0

V I

J ~zw0 >

I a. 0

Iw cI2

0 I-- cc0o.L 0...,

OD0N Z

> <0

ILI0a 2i -m >- w

ani-UanWuP w CD

x 0- x -< 0 ILO U w k.

z cc 133

0 0o~

C

S 0=

CL3

oco

- 0~0-0

CD

3 CL~0

0.

x Z

z -

I- E.

F-

1111..- ,'"

0 a-

CM

0 1

I,

134U

U-

hil II I hi l 1II34L

>109

0 RADAR ECHOES

0 NO RADAR ECHOES

106

: log

'• !'°P i l l I •I .,,i I i i

' oRi I i I 1 1 1 1

%_

w !I

1088

io6 i I i I i I I I i I I •6 10711 ~ li i-

" o -] ! ! I I I I I l! I ! 3 HOUR -

S0 0 1 1 1 2 2 3 4 5 5 6 7 7 K-INDEX

,• '0 0 PASS

Oa t . c ' NUMBER

S1 2 3 4 5 6 7 8 9 10 II 12 13 14 SEQUENCE

NUMBER=

S~FIGURE 6.6 AVERAGE PROPERTIES OF PRECIPITATING ELECTRONS versus SATELLITE

S~PASS SEQUENCE NUMBER ARRANGED ACCORDING TO ASCENDING

• THREE-HOUR MAGNETIC K-INDEX

1135

6 0f~5

~ ~• ~ ->

SI

I,.°

110

I- C- 0.

10

to

.; W

0

0. Z4 I I I I '

10

0

.w -i

Iu0.1

10

Z V

"'o • I I I I I I I j

3UOU0.

0 0 1 1 1 2 2 3 4 5 1 6 7 7 3-HOURIn m

o to CD C.) N 0 w "r cm C9 NUMBER

1 2 3 4 5 6 7 8 9 10 11 12 13 14 SEQUERNC

FIGURE 6.7 RATIO OF THE AVERAGE PROPERTIES OF PRECIPITATING ELECTRONSIN REGIONS WHERE RADAR AURORA WAS OBSERVED, TO THOSEPROPERTIES IN REGIONS WHERE NO RADAR AURORA WAS OBSERVED,

versus SATELLITE-PASS SEQUENCE NUMBER ARRANGED ACCORDING TO •ASCENDING THREE-HOUR MA.GNETIC K-NDEX

136<

35

7. CONCLUSIONS

This section summarizes what this research effort has revealed.

Comparison of the location of peak precipitation energy and number

density of electrons as given in Figures 4.C.1 through 4.C.14, with simi-

lar quantities for protons in Figures 4.D.1 through 4.D.14 shows that,

with the exception of Passes A2964. A2995, and A2979, protons precipitate

to the south of electrons. This was the case for the eleven passes at

lower K indices without exception. For the three passes that took place

during high magnetic activity, the satellite precipitation data were not

provided far enough south to verify that this behavior was also present

during these three passes.

The data on these 14 satellite passes indicate that the electron

density in the nighttime E-region that is necessary for radar scatteringis probably produced by electron precipitation.

With the exception of the three passes that took place during very

intense magnetic activity, radar aurora was never observed to be collo-

cated with regions of peaR proton-number flux or proton-energy-flux pre-

cipitation. On the three passes that took pLace during high magnetic

activity, as we indicated above, satellite data at more southerly lati-

tudes were not available. so that it is not knonn whether the proton data

that were obtained for these three passes contain the most significant

proton precipitation. Thus, it cannot be said with certainty thal the

coincidences between proton precipitation and radar aurora on these three

passes contradicts the conclusions that have been drawn from the eleven

other passes. We believe that this anticorrelation between protov pre-

cipitation and radar aurora is significant.

137

Attempts to find some deterministic relationship between the de-

tailed characteristics of the precipitating electrons and the regions

from which radar aurora is present have not produced any unique relation-

ship. There appear to be some statistical trends, though these are

subtle. These trends are as follows:

(1) There may be a tendency for regions of locally more

intense electron-energy flux to anticorrelate with

radar aurora, though this is not an absolute tendency.

(2) There may be a slight tendency for the average particle

number flux--TSTh¶--to be less in regions of radar scattering

than in other regions where measurable electron precipi-

tation is taking place.

(3) There is a high statistical likelihood that the average

electron precipitation energy flux (ELE) is less in

regions from which radar echoes are obtained.

(4) The average energy of the precipitating electron is, on

the average, less in regions of radar aurora.

r 138!Uo I

8. SUGGESTIONS FOR FURTHER WORK

Radar data for 54 additional satellite passes have been reduced at

SRI. Comparative studies must await delivery of Lockheed particle data

which appears unlikely to occur due to Lockheed funding limitations.

Comparison between these additional 54 satellite passes would certainly

be helpful in understanding aurora phenomena, in that the statistics of

the phenomena will be greatly improved. Data from input-output experi-

ments are still very rare. J

The fourteen satellite passes that ha',e been reduced have not been

as thoroughly analyzed as is possible. The radar data that are presented

in this report were reduced for the time period corresponding to the pre-

dicted satellite ephemeris data. As a result, the simultaneity of the

radar data that we have analyzed, and the satellite pass data, is in some

cases very poor. This lack of good simultaneity can be corrected for

much of the data.

Many data correlations involving radar auroral locations, radar

auroral signal strength, precipitation-particle energies, particle fluxes,

etc., that could be performed have not received attention. In order to

do an adequate job on future correlations, the radar data must be digi-

tized and then recompared with the digitized satellite data. We beLieve

such 2 rerun atA a second analysis is mozo- important than would be addi-

tional work on the 54 satellite passes for which Lockheed has not supplied

us with precipitation data. For iur comparison work, the data that

In order to obtain the Lockheed data for 14 passes analyzed here it was

necessary for SRI to fund Lockheed directly.

139

Lockheed delivers to us in the future should give, if possible, the in-

dividual electron channeltron counter rates rather than the summary value

EMN. This would help us more accurately determine from the satellite

data the altitudes at which the electron-precipitation energy is converted

to ambient E-region ionization. For our comparison work we need to extend

the work of Rees 6 ' 7 and Chamberlain9 to better meet the needs of this

progran. In particular, we need to compute depositioil for particle spec-

tra and pitch-angle distributions that are more 1'., those that the

satellites actually measure. Furthermore, we must look into the effects

of proton-charge-exchange scattering in order to do 3 better job of making

spatial comparisons between aurora and proton precipitation.

140

I!_

REFERENCES

1. R. L. Leadabrand and J. C. Hodges, "Correlation of Radar Echoes

From the Aurora with Satellite-Measures Auroral Particle Precipita-tion," J. Geophys. Res., Vol. 72, No. 21, pp. 5,411-5317 (1 November

1967).

2. W. G. Chesnut, J. C. Hodges, and R. L. Leadabrand, "Auroral Back-scatter Wavelength Dependence Studies," Final Report, Contract AF30 (602)-37345 SRI Project 5535, Stanford Research Institute, MenloPark, California (June 1968).

3. T. H~agfors, R. G. Johnson, and R. A. Powers. "Simultaneous Observa-tion of Proton Precipitation and Auroral Radar Echoes," J. Geophys.Res. Vol. 76, No. 25, pp. 6093-6098 (1 September 1971).

4. J.E. Evans and R. G. Johnsoa, "Auroral Input-Output Measurements,"

Contract Nonr-3398(00), Task No. NR 087-135, Lockheed Missiles andSpace Company, Palo Alto, California.

S 5. G. B. Shook. "A Coordinated Experiment on Radar Clutter and AuroralParticle Fluxes," Final Report, Space Sciences Laboratory, Contract

N00014-69-C-0207. Task No. NR 087-170, Lockheed Missiles and Space

Company, Palo Alto, California (30 April 1970).

6. M. Rees. "Note on the Penetration of Energetic Electrons into the

Earth's Atmosphere," Planet Space Sci., Vol. 12, pp. 722-725 (1964).

7. M. Rees "Auroral Ionization and Excitation by Incident EnergeticElectrons," Planet Space Sci., Vol. 11, pp. 1209-1218 (1963).

8. I. G. Poppoff and R. C. Whitten. DASA Reaction Rate Handbook, Chapt.8, "Reaction Rate Data From Atmospheric Measurements," DASA-1948

(October 1967).

9. J. W. Chamberlain, Physics of the Aurora and Airglow (AcademicPress, New York, N. Y.. 1961).

10. E. M. Wescott, J. D. Stolarik, and J. P. Heppner, "Electric Fieldsin the Vicinity of Auroral Forms from Motions of Barium Vapor

Releases, J. Geophys. Res.. Vol. 74, No. 14, pp. 3469-3487 (July1, 1969).

141

S11. T. L. Aggson, "Probe Measurement of Electric Fields in Space." in

Atmospheric Emissions, B. M. McCormac and A. Omholt, Eds., p. 305

(Van Nostrand Reinhold Co., New York, N.Y., 1969).

14

S~142

dna 2825f - dtic00 dna 2825f q4 final report december 1971 correlation of radar echoes from the...

Documents